fix(security): scope cert-sign participant listing to caller's team (GHSA-h2rx-xrhc-5q72)

2026-09-03 05:10:16 +03:00 · 2026-06-08 00:00:04 +01:00
2932 changed files with 106885 additions and 342553 deletions
@@ -1,97 +0,0 @@
---
-name: feature-walkthrough
-description: >-
-  Explain the full logic and process of the current branch end-to-end so someone
-  with no prior knowledge of the task can understand, review, and reproduce it.
-  Scopes the change from the branch diff, traces the flow across every layer it
-  touches (frontend tool/hook/component, Java controller/service/endpoint, Python
-  engine, config, i18n, tests), and produces a self-contained walkthrough document
-  with Mermaid diagrams (sequence/flow/architecture), annotated file map with
-  clickable references, before/after behavior, screenshots where a UI is involved,
-  a "try it locally" section, and edge cases/risks. Use when asked for a feature or
-  branch walkthrough, "explain what this branch does", a design/logic writeup, PR
-  reviewer onboarding, or a hand-off doc. Pass --html to also emit a rendered HTML
-  version; --no-screens to skip screenshots.
-argument-hint: "[branch-or-area] [--html] [--no-screens]"
-allowed-tools: Read, Write, Edit, Glob, Grep, Bash
---
-
-# Feature / Branch Walkthrough
-
-Turn the current branch into a walkthrough a newcomer can follow. Audience:
-**someone who has never seen this task**. Explain the *why*, the *flow*, and *how to
-try it* - not just a diff summary.
-
-`$ARGUMENTS` may name a branch or area to focus on; default is the current branch
-vs `main`. Flags: `--html` (also emit a rendered HTML twin), `--no-screens`.
-
-## Process
-
-### 1. Scope the change
- `git log --oneline main..HEAD` and `git diff --stat main...HEAD` for the shape.
- Read the PR description / commit messages for stated intent. Do **not** invent
-  history or motivation that isn't evidenced (state current behavior in present tense).
- Classify touched files by layer:
-  - **Frontend**: tools (`frontend/editor/src/core/components/tools/*` or `.../core/tools/*`),
-    hooks (`core/hooks/tools/*`, `useToolOperation`), contexts, routes, i18n
-    (`public/locales/en-US`).
-  - **Java backend**: controllers (`.../controller/api/...`), services, models, config.
-  - **Engine**: `engine/src/stirling/{agents,contracts,api,services}`.
-  - **Config / build / docker / tests.**
-
-### 2. Trace the flow end-to-end
-Follow one real path from user action to result. For a typical PDF tool that's:
-UI control → `useToolOperation` hook → `POST /api/v1/...` → Spring controller →
-service (PDFBox / LibreOffice / engine call) → response → review panel → download.
-Read the actual files so the narrative is true to the code, and collect the exact
-file:line anchors you'll cite.
-
-### 3. Draw the diagrams (Mermaid)
-Pick what fits; usually 2-3 of:
- **Sequence diagram** - request/response across frontend → backend → engine.
- **Flowchart** - the core decision/branching logic of the feature.
- **Architecture/component** - new pieces and how they wire to existing ones.
- **State** - if the feature has modes/steps.
-Keep nodes labeled in plain language. Validate the Mermaid parses before shipping.
-
-### 4. Screenshots (unless --no-screens)
-If a UI is involved, capture key states with the stubbed Playwright harness
-(see the **ui-walkthrough** skill and `files-page-screenshots.spec.ts` for the
-pattern) or, for before/after, capture `main` then the branch. Drop PNGs in
-`walkthrough/<feature>/` and reference them from the doc. For backend-only
-changes, show request/response examples (curl + JSON) instead.
-
-### 5. Write the walkthrough
-Create `walkthrough/<feature>/FEATURE-WALKTHROUGH.md` with:
-1. **TL;DR** - what the branch does and who it's for, in 3-4 sentences.
-2. **Problem & approach** - what wasn't possible before; the chosen solution.
-3. **Architecture diagram** + 1-paragraph orientation.
-4. **End-to-end flow** - the sequence diagram + a numbered walk of each step,
-   each citing the real file (clickable `path:line`).
-5. **Key files** - annotated map (path → one line on its role).
-6. **Logic deep-dive** - the flowchart + prose for the non-obvious decisions.
-7. **Behavior** - before vs after; screenshots or request/response examples.
-8. **Try it locally** - exact steps (`task dev` / `task dev:all`, the route to
-   open or the curl to run, any env like `DOCKER_ENABLE_SECURITY` or a test
-   license key). Make it copy-pasteable.
-9. **Edge cases, risks, follow-ups** - what's untested, known limits, gotchas.
-
-Markdown is the primary deliverable - it renders with diagrams in GitHub PRs and
-IDEs, no build step, ideal for review.
-
-### 6. If `--html`
-Also emit `walkthrough/<feature>/walkthrough.html`: the same content with Mermaid
-rendered via `mermaid.initialize({startOnLoad:true})` (script from CDN; note in
-the file that rendering diagrams needs network, the `.md` is the offline copy) and
-screenshots inline. Keep it self-contained otherwise.
-
-### 7. Deliver
-Give the doc path and a short chat summary. Offer to `SendUserFile` it.
-
-## Principles
- **True to the code.** Every claim traces to a file you read; cite `path:line`.
-  No fabricated migration/version history.
- **Newcomer-first.** Define repo-specific terms (FileContext, `useToolOperation`,
-  the `@app/*` layer cascade, stubbed vs live tests) on first use.
- **Show, don't assert.** Prefer a diagram + a real example over adjectives.
- Don't commit the `walkthrough/` output unless asked.
@@ -1,122 +0,0 @@
---
-name: ui-before-after
-description: >-
-  Analyse a branch or PR and automatically capture before/after screenshots of
-  every UI surface its changes touch, then pixel-diff the pairs to surface what
-  actually changed and assemble PR-ready before/after montage images. Generic and
-  diff-driven: it derives the capture targets from the diff (changed tools/routes →
-  URLs) instead of hand-listing screens, captures "before" from the base branch and
-  "after" from the head, then keeps only the views that visually differ. Each
-  comparison is auto-cropped to the region that actually changed (the bounding box of
-  differing pixels), falling back to the full page only when the change spans most of
-  it. Use for before/after shots, a visual diff of a branch/PR, "screenshots for the
-  PR description", "show what changed in the UI", or a side-by-side of UI changes.
-  Takes a PR number/URL (resolved via gh) or a branch; defaults to the current branch
-  vs its base. Flags: --scope <selector>, --base <ref|merge-base>, --theme
-  light|dark|both, --all (capture every route, not just changed), --no-autocrop,
-  --pagewide <n>, --threshold <n>.
-argument-hint: "[PR# | PR-url | branch] [--scope <sel>] [--base <ref>] [--theme both] [--all] [--no-autocrop]"
-allowed-tools: Read, Write, Edit, Glob, Grep, Bash
---
-
-# UI Before / After (generic visual diff)
-
-Point it at a branch or PR; it figures out which UI changed, screenshots every
-affected surface **before** (base) and **after** (head), pixel-diffs the pairs, and
-montages the ones that actually changed into images for the PR description.
-
-`$ARGUMENTS`: a PR number/URL, a branch, or nothing (current branch vs base).
-By default it captures the full viewport and auto-crops each comparison to the region
-that changed. Flags: `--scope <css>` (narrow the *capture* to a container, e.g.
-`[data-sidebar="tool-panel"]`, when you already know where the change is),
-`--no-autocrop` (keep full frames), `--pagewide <fraction>` (above this share of the
-page, skip cropping; default 0.6), `--base <ref|merge-base>`,
-`--theme light|dark|both`, `--all` (walk every route, not just changed),
-`--threshold <fraction>` (diff sensitivity, default 0.001).
-
-Shares the capture harness with **ui-walkthrough** - read its SKILL.md for the
-stubbed-Playwright setup, worktree node_modules + `generate-icons`, the
-stale-`:5173` gotcha, and the dark-mode init-script. Bundled helpers:
-[capture-spec.template.ts](capture-spec.template.ts), [diff-shots.mjs](diff-shots.mjs),
-[montage-template.html](montage-template.html), [shoot-sections.mjs](shoot-sections.mjs).
-
-## Process
-
-### 1. Resolve target + base
-```
-gh pr view <pr> --json number,title,headRefName,baseRefName,url,files   # PR
-# or branch: base = merge-base(main, HEAD); head = HEAD
-gh pr diff <pr> --name-only        # or: git diff --name-only <base>...HEAD
-```
-
-### 2. Derive capture targets from the diff (the "analyse" step - no hand-listing)
-Map changed frontend files to URLs generically:
- **Tools**: a changed `components/tools/<toolDir>/…` or `hooks/tools/<tool>/…` →
-  toolId → URL via the repo's own rule `getToolUrlPath` in
-  [toolsTaxonomy.ts:200](frontend/editor/src/core/data/toolsTaxonomy.ts): `/` + the
-  id kebab-cased (`addPageNumbers` → `/add-page-numbers`).
- **Pages/routes**: changed `filesPage/*` → `/files`, etc.
- `--all`: enumerate every tool in the registry instead of just changed ones.
-Write `frontend/editor/screenshots/ui-diff/targets.json` =
-`[{ "id":"compress", "url":"/compress", "name":"Compress" }]`. This is what makes
-it generic - the spec never names a tool.
-
-### 3. Capture AFTER (head) then BEFORE (base)
-Copy [capture-spec.template.ts](capture-spec.template.ts) →
-`src/core/tests/stubbed/ui-before-after.spec.ts` (it loops `targets.json`, seeds a
-sample PDF so file-dependent panels render, navigates to each URL, and screenshots
-the full viewport - or the `--scope` container if given). Ensure the harness is ready
-(node_modules + icons).
-```
-# after = current head
-cd frontend/editor && PR_SHOT_SIDE=after PR_SHOT_THEME=light \
-  npx playwright test --project=stubbed ui-before-after.spec.ts
-# before = base, in an isolated worktree (copy the spec + targets.json in)
-git worktree add ../ba-base origin/<baseRefName>        # or the merge-base
-#   set up its frontend, copy spec + screenshots/ui-diff/targets.json across, then:
-cd ../ba-base/frontend/editor && PR_SHOT_SIDE=before PR_SHOT_THEME=light \
-  npx playwright test --project=stubbed ui-before-after.spec.ts
-#   copy its screenshots/ui-diff/before/ back next to after/. Repeat with
-#   PR_SHOT_THEME=dark if --theme includes dark. Remove worktree when done.
-```
-
-### 4. Auto-diff (surface what changed)
-```
-cd frontend/editor && node <skill>/diff-shots.mjs \
-  screenshots/ui-diff/before screenshots/ui-diff/after screenshots/ui-diff
-```
-Produces `diff-report.json` classifying each view `unchanged | changed | added |
-removed`. For each changed view it computes the bounding box of differing pixels and
-writes cropped `__before_crop.png` / `__after_crop.png` / `__diff.png` to that region
-(+ padding) - **unless** the change covers more than `--pagewide` of the frame, where
-it keeps the full frame (`pageWide:true`). Drop `unchanged` - that's the noise the
-user doesn't want.
-
-### 5. Montage the changes
-Build the manifest from the non-unchanged entries (group by tab/tool; each becomes a
-state row with before/after). For changed views use the cropped `cropBefore` /
-`cropAfter` from `diff-report.json` (tight on the affected region; full frame when
-`pageWide`); `added`/`removed` render the "not present" placeholder. Fill
-[montage-template.html](montage-template.html) (replace the `window.__BA__` data
-block; base64-inline the PNGs for portability), then render one PNG per section with
-[shoot-sections.mjs](shoot-sections.mjs). Optionally include the `__diff.png` overlay
-as a third column.
-
-### 6. Deliver
-Output the `montage_<tab>.png` files + a short summary (N changed / added / removed,
-M unchanged skipped) and a paste-ready Markdown block. GitHub has no PR-body image
-API, so tell the user to drag the PNGs into the description. Do **not** post to the
-PR.
-
-## Gotchas
- Two installs (base worktree + head); junction main's node_modules only if its deps
-  match that ref, else `npm ci` (see ui-walkthrough's stale-dep note).
- A view that errors on one side (refactored/removed) → that side is missing; the
-  diff marks it added/removed rather than failing the run.
- Pixel diff needs equal dimensions, so capture at a fixed viewport (the template
-  does); a view whose size changed is reported as "changed (dimensions differ)",
-  uncropped.
- Auto-crop uses a single bounding box, so two far-apart changes give one large crop
-  (or trip `--pagewide`); narrow with `--scope` if that happens.
- `getToolUrlPath` is the source of truth for tool URLs - use it, don't guess slugs.
- Don't commit `screenshots/`, the throwaway spec, or the base worktree.
@@ -1,67 +0,0 @@
-// Generic before/after capturer. NOT app-specific: it walks a targets.json that
-// the ui-before-after skill generates from the branch/PR diff, so nothing here is
-// hand-listed. Copy to src/core/tests/stubbed/ui-before-after.spec.ts, then run
-// once per (side, theme):
-//   PR_SHOT_SIDE=after  PR_SHOT_THEME=light \
-//     npx playwright test --project=stubbed ui-before-after.spec.ts
-//
-// targets.json shape: [{ "id":"compress", "url":"/compress", "name":"Compress",
-//                        "needsFile": true }]
-import { test } from "@app/tests/helpers/stub-test-base";
-import type { Page } from "@playwright/test";
-import fs from "node:fs";
-import path from "node:path";
-
-const SIDE = process.env.PR_SHOT_SIDE ?? "after";
-const THEME = process.env.PR_SHOT_THEME ?? "light";
-// Capture the full viewport by default so the affected region is in frame
-// wherever it is; diff-shots.mjs crops each comparison to what actually changed.
-// Set PR_SHOT_SCOPE to a selector to narrow the capture to one container.
-const SCOPE = process.env.PR_SHOT_SCOPE ?? "";
-const ROOT = path.resolve(process.cwd(), "screenshots", "ui-diff");
-const OUT = path.join(ROOT, SIDE);
-// A tiny sample PDF so file-dependent tool panels render. Point at a real fixture.
-const SAMPLE_PDF = process.env.PR_SHOT_SAMPLE ?? "src/core/tests/test-fixtures/sample.pdf";
-
-type Target = { id: string; url: string; name?: string; needsFile?: boolean };
-const targets: Target[] = JSON.parse(fs.readFileSync(path.join(ROOT, "targets.json"), "utf-8"));
-
-test.use({ autoGoto: false, viewport: { width: 1600, height: 900 }, seedJwt: true });
-
-async function applyTheme(page: Page): Promise<void> {
-  if (THEME !== "dark") return;
-  await page.addInitScript(() => {
-    localStorage.setItem("mantine-color-scheme", "dark");
-    localStorage.setItem("mantine-color-scheme-value", "dark");
-  });
-  await page.emulateMedia({ colorScheme: "dark" });
-}
-
-async function seedFile(page: Page): Promise<void> {
-  if (!fs.existsSync(SAMPLE_PDF)) return;
-  await page.goto("/", { waitUntil: "domcontentloaded" });
-  await page.getByTestId("files-button").click().catch(() => {});
-  await page.locator('[data-testid="file-input"]').setInputFiles(SAMPLE_PDF).catch(() => {});
-  await page.locator(".file-sidebar-file-item").first().isVisible({ timeout: 8_000 }).catch(() => {});
-}
-
-for (const t of targets) {
-  // One test per target so a single failure doesn't drop the rest.
-  test(`${SIDE}/${THEME} ${t.id}`, async ({ page }) => {
-    fs.mkdirSync(OUT, { recursive: true });
-    await applyTheme(page);
-    if (t.needsFile !== false) await seedFile(page);
-    await page.goto(t.url, { waitUntil: "domcontentloaded" });
-    await page.waitForTimeout(400); // settle Mantine portals/transitions
-    const shot = path.join(OUT, `${t.id}__${THEME}.png`);
-    if (SCOPE) {
-      const scope = page.locator(SCOPE).first();
-      if (await scope.isVisible({ timeout: 8_000 }).catch(() => false)) {
-        await scope.screenshot({ path: shot });
-        return;
-      }
-    }
-    // Full viewport (fixed size → stable dimensions for pixel diffing).
-    await page.screenshot({ path: shot });
-  });
-}
@@ -1,106 +0,0 @@
-// Auto-diff before/ vs after/ screenshots, classify each as
-// unchanged | changed | added | removed, and CROP each changed pair to the
-// affected region (bounding box of differing pixels + padding) - unless the
-// change spans most of the page, in which case the full frame is kept.
-// Run from frontend/editor (so deps resolve):
-//   node <skill>/diff-shots.mjs <beforeDir> <afterDir> [outDir]
-// Env:
-//   DIFF_THRESHOLD  min fraction of differing pixels to count as changed (default 0.001)
-//   DIFF_PAD        padding px around the affected region (default 24)
-//   DIFF_PAGEWIDE   if affected bbox area / image area exceeds this, keep full frame (default 0.6)
-import fs from "node:fs";
-import path from "node:path";
-import { createRequire } from "node:module";
-
-const require = createRequire(path.join(process.cwd(), "noop.js"));
-const pm = require("pixelmatch");
-const pixelmatch = pm.default || pm;
-const { PNG } = require("pngjs");
-
-const beforeDir = path.resolve(process.argv[2]);
-const afterDir = path.resolve(process.argv[3]);
-const outDir = path.resolve(process.argv[4] || afterDir);
-const THRESHOLD = Number(process.env.DIFF_THRESHOLD ?? "0.001");
-const PAD = Number(process.env.DIFF_PAD ?? "24");
-const PAGEWIDE = Number(process.env.DIFF_PAGEWIDE ?? "0.6");
-
-const read = (p) => PNG.sync.read(fs.readFileSync(p));
-const isShot = (f) => f.endsWith(".png") && !/__(diff|before_crop|after_crop)\.png$/.test(f);
-const list = (d) => (fs.existsSync(d) ? fs.readdirSync(d).filter(isShot) : []);
-const names = [...new Set([...list(beforeDir), ...list(afterDir)])].sort();
-fs.mkdirSync(outDir, { recursive: true });
-
-function cropPNG(src, x, y, w, h) {
-  const out = new PNG({ width: w, height: h });
-  PNG.bitblt(src, out, x, y, w, h, 0, 0);
-  return out;
-}
-const writePNG = (p, png) => fs.writeFileSync(p, PNG.sync.write(png));
-
-// Bounding box of differing pixels using a diff mask (alpha>0 where changed).
-function changedBBox(before, after, w, h) {
-  const mask = new PNG({ width: w, height: h });
-  pixelmatch(before.data, after.data, mask.data, w, h, { threshold: 0.1, diffMask: true });
-  let minX = w, minY = h, maxX = -1, maxY = -1, count = 0;
-  for (let y = 0; y < h; y++) {
-    for (let x = 0; x < w; x++) {
-      if (mask.data[(y * w + x) * 4 + 3] > 0) {
-        count++;
-        if (x < minX) minX = x; if (x > maxX) maxX = x;
-        if (y < minY) minY = y; if (y > maxY) maxY = y;
-      }
-    }
-  }
-  return maxX < 0 ? null : { minX, minY, maxX, maxY, count };
-}
-
-const report = [];
-for (const name of names) {
-  const id = name.replace(/\.png$/, "");
-  const bp = path.join(beforeDir, name), ap = path.join(afterDir, name);
-  const hasB = fs.existsSync(bp), hasA = fs.existsSync(ap);
-  if (hasB && !hasA) { report.push({ id, status: "removed", before: bp }); continue; }
-  if (!hasB && hasA) { report.push({ id, status: "added", after: ap }); continue; }
-
-  const before = read(bp), after = read(ap);
-  if (before.width !== after.width || before.height !== after.height) {
-    report.push({ id, status: "changed", note: "dimensions differ", before: bp, after: ap });
-    continue;
-  }
-  const w = after.width, h = after.height;
-  const overlay = new PNG({ width: w, height: h });
-  const px = pixelmatch(before.data, after.data, overlay.data, w, h, { threshold: 0.1 });
-  const ratio = px / (w * h);
-  if (ratio <= THRESHOLD) { report.push({ id, status: "unchanged", ratio: Number(ratio.toFixed(5)), before: bp, after: ap }); continue; }
-
-  const box = changedBBox(before, after, w, h);
-  // Pad + clamp the affected region.
-  const x = Math.max(0, box.minX - PAD), y = Math.max(0, box.minY - PAD);
-  const x2 = Math.min(w, box.maxX + 1 + PAD), y2 = Math.min(h, box.maxY + 1 + PAD);
-  const bw = x2 - x, bh = y2 - y;
-  const pageWide = (bw * bh) / (w * h) > PAGEWIDE;
-
-  const entry = { id, status: "changed", ratio: Number(ratio.toFixed(5)), before: bp, after: ap, pageWide };
-  if (pageWide) {
-    // Change spans most of the page - keep the full frame, full overlay.
-    const dp = path.join(outDir, `${id}__diff.png`); writePNG(dp, overlay);
-    entry.diff = dp;
-  } else {
-    entry.bbox = { x, y, w: bw, h: bh };
-    const cb = path.join(outDir, `${id}__before_crop.png`); writePNG(cb, cropPNG(before, x, y, bw, bh));
-    const ca = path.join(outDir, `${id}__after_crop.png`); writePNG(ca, cropPNG(after, x, y, bw, bh));
-    const dp = path.join(outDir, `${id}__diff.png`); writePNG(dp, cropPNG(overlay, x, y, bw, bh));
-    entry.cropBefore = cb; entry.cropAfter = ca; entry.diff = dp;
-  }
-  report.push(entry);
-}
-
-fs.writeFileSync(path.join(outDir, "diff-report.json"), JSON.stringify(report, null, 2));
-const changed = report.filter((r) => r.status !== "unchanged");
-console.log(`diffed ${report.length} view(s): ${changed.length} changed/added/removed, ${report.length - changed.length} unchanged`);
-for (const r of changed) {
-  const tail = r.status !== "changed" ? ""
-    : r.pageWide ? " (page-wide → full frame)"
-    : ` (${(r.ratio * 100).toFixed(2)}%, cropped to ${r.bbox.w}×${r.bbox.h})`;
-  console.log(`  ${r.status.padEnd(9)} ${r.id}${tail}${r.note ? " - " + r.note : ""}`);
-}
@@ -1,48 +0,0 @@
-"""Build EXAMPLE.html from montage-template.html using REAL files-page shots as
-stand-in before/after pairs (layout demo, not an actual PR diff). Inlines PNGs as
-data URIs so the HTML is portable. Run: python make_example.py"""
-import base64
-import json
-import pathlib
-import re
-
-HERE = pathlib.Path(__file__).parent
-SHOTS = pathlib.Path(
-    r"C:\Users\systo\git\Stirling-PDFNew\.claude\worktrees\kind-faraday-522a30"
-    r"\frontend\editor\screenshots\files-page"
-)
-
-
-def uri(fname):
-    p = SHOTS / fname
-    return "data:image/png;base64," + base64.b64encode(p.read_bytes()).decode() if p.exists() else None
-
-
-data = {
-    "pr": "DEMO",
-    "title": "EXAMPLE — before/after montage (layout demo, real Files-page shots; not a real PR diff)",
-    "base": "main", "head": "demo-branch",
-    "cropSelector": "[data-sidebar=\"tool-panel\"]  (real runs crop to the side; these demo shots are full-page)",
-    "tabs": [
-        {"id": "files", "title": "Files page", "ctx": "Each row = one flow state; left = base branch, right = this PR.",
-         "states": [
-             {"name": "Empty folder", "before": uri("01_empty_state_ctas.png"), "after": uri("02_empty_state_storage_off.png")},
-             {"name": "Files + details panel", "before": uri("03_subtoolbar_with_files.png"), "after": uri("06_details_panel_save_to_server.png")},
-             {"name": "Delete folder confirm", "before": None, "after": uri("19_delete_folder_dialog.png"), "note": "New in this PR"},
-         ]},
-        {"id": "move", "title": "Move-to-folder dialog",
-         "states": [
-             {"name": "Dialog opened", "before": uri("07_move_dialog_collapsed.png"), "after": uri("08_move_dialog_create_folder_expanded.png")},
-             {"name": "After folder created", "before": None, "after": uri("08b_move_dialog_after_create_folder.png"), "note": "New flow"},
-         ]},
-    ],
-}
-
-tpl = (HERE / "montage-template.html").read_text(encoding="utf-8")
-out = re.sub(
-    r"/\*__DATA__\*/.*?/\*__END__\*/",
-    lambda _m: "/*__DATA__*/" + json.dumps(data) + "/*__END__*/",
-    tpl, count=1, flags=re.S,
-)
-(HERE / "EXAMPLE.html").write_text(out, encoding="utf-8")
-print("wrote", HERE / "EXAMPLE.html", "(", (HERE / "EXAMPLE.html").stat().st_size // 1024, "KB )")
@@ -1,106 +0,0 @@
-<!doctype html>
-<!--
-  Before/After montage for a PR description. The ui-before-after skill replaces
-  the JSON in the window.__BA__ data block below with the captured manifest, then
-  screenshots each .tab-section (id="section-<tabId>") into a PNG to drag into the
-  PR description. Self-contained; images may be relative paths or data URIs.
-
-  Data shape:
-  {
-    "pr":"6552","title":"...","base":"main","head":"feat/x",
-    "cropSelector":"[data-sidebar=\"tool-panel\"]",
-    "tabs":[
-      { "id":"sign","title":"Sign tool","states":[
-        {"name":"Initial","before":"before/sign__initial.png","after":"after/sign__initial.png"},
-        {"name":"Cert selected","before":null,"after":"after/sign__cert.png","note":"New in this PR"}
-      ]}
-    ]
-  }
-->
-<html lang="en">
-<head>
-<meta charset="utf-8" />
-<title>Before / After</title>
-<style>
-  :root { --bg:#ffffff; --ink:#0b0c0e; --muted:#6b7280; --line:#e5e7eb;
-          --before:#6b7280; --after:#1f883d; --frame:#f3f4f6; --note:#b45309; }
-  * { box-sizing: border-box; }
-  body { margin:0; background:var(--bg); color:var(--ink);
-         font:14px/1.5 -apple-system,"Segoe UI",Roboto,system-ui,sans-serif; }
-  .wrap { max-width:1100px; margin:0 auto; padding:24px; }
-  .doc-head { margin-bottom:8px; }
-  .doc-head h1 { font-size:18px; margin:0 0 2px; }
-  .doc-head .sub { color:var(--muted); font-size:12.5px; }
-  .legend { display:flex; gap:14px; align-items:center; margin:10px 0 4px; font-size:12px; color:var(--muted); }
-  .chip { font-size:10px; font-weight:700; letter-spacing:.04em; text-transform:uppercase;
-          padding:2px 8px; border-radius:999px; color:#fff; }
-  .chip.before { background:var(--before); } .chip.after { background:var(--after); }
-
-  .tab-section { border:1px solid var(--line); border-radius:14px; padding:18px 18px 8px;
-                 margin:18px 0; background:var(--bg); }
-  .tab-section > h2 { font-size:16px; margin:0 0 2px; }
-  .tab-section > .ctx { color:var(--muted); font-size:12px; margin-bottom:14px; }
-  .state { margin-bottom:18px; }
-  .state .name { font-weight:600; font-size:13.5px; margin-bottom:8px; display:flex; gap:8px; align-items:center; }
-  .state .name .note { font-weight:500; color:var(--note); font-size:12px; }
-  .pair { display:grid; grid-template-columns:1fr 1fr; gap:14px; align-items:start; }
-  .cell { border:1px solid var(--line); border-radius:10px; overflow:hidden; background:var(--frame); }
-  .cell .cap { display:flex; align-items:center; gap:8px; padding:7px 10px; border-bottom:1px solid var(--line);
-               background:var(--bg); }
-  .cell .cap .meta { color:var(--muted); font-size:11px; }
-  .cell img { display:block; width:100%; height:auto; background:#fff; }
-  .cell.empty .ph { display:flex; align-items:center; justify-content:center; height:160px; color:var(--muted);
-                    font-size:12.5px; text-align:center; padding:0 16px; }
-  .single .pair { grid-template-columns:1fr; }
-  .empty-doc { color:var(--muted); padding:40px; text-align:center; }
-  @media (max-width:760px){ .pair{ grid-template-columns:1fr; } }
-</style>
-</head>
-<body>
-<div class="wrap" id="root"></div>
-
-<script id="data">
-window.__BA__ = /*__DATA__*/{"pr":"","title":"No data","base":"","head":"","cropSelector":"","tabs":[]}/*__END__*/;
-</script>
-<script>
-(function(){
-  var D = window.__BA__ || { tabs: [] };
-  var root = document.getElementById("root");
-  function el(html){ var t=document.createElement("template"); t.innerHTML=html.trim(); return t.content.firstChild; }
-  function esc(s){ return (s==null?"":String(s)).replace(/[&<>]/g, function(c){return {"&":"&amp;","<":"&lt;",">":"&gt;"}[c];}); }
-
-  function cell(kind, src){
-    if (src) {
-      return '<div class="cell"><div class="cap"><span class="chip '+kind+'">'+kind+'</span></div>'+
-             '<img src="'+esc(src)+'" alt="'+kind+'"/></div>';
-    }
-    return '<div class="cell empty"><div class="cap"><span class="chip '+kind+'">'+kind+'</span>'+
-           '<span class="meta">not present</span></div><div class="ph">No '+kind+' screenshot for this state</div></div>';
-  }
-
-  var head = '<div class="doc-head"><h1>'+esc(D.title || ("PR #"+D.pr))+'</h1>'+
-    '<div class="sub">Before / after &nbsp;·&nbsp; base <code>'+esc(D.base)+'</code> → head <code>'+esc(D.head)+'</code>'+
-    (D.cropSelector ? ' &nbsp;·&nbsp; cropped to <code>'+esc(D.cropSelector)+'</code>' : '')+'</div></div>'+
-    '<div class="legend"><span class="chip before">Before</span> base branch'+
-    '<span class="chip after">After</span> this PR</div>';
-  root.appendChild(el('<div>'+head+'</div>'));
-
-  if (!D.tabs || !D.tabs.length){ root.appendChild(el('<div class="empty-doc">No tabs captured yet.</div>')); return; }
-
-  D.tabs.forEach(function(tab){
-    var states = (tab.states||[]).map(function(s){
-      var onlyOne = (!s.before || !s.after);
-      return '<div class="state'+(onlyOne?' ':'')+'">'+
-        '<div class="name">'+esc(s.name)+(s.note?'<span class="note">'+esc(s.note)+'</span>':'')+'</div>'+
-        '<div class="pair">'+cell("before", s.before)+cell("after", s.after)+'</div></div>';
-    }).join("");
-    var sec = '<section class="tab-section" id="section-'+esc(tab.id)+'">'+
-      '<h2>'+esc(tab.title)+'</h2>'+
-      (tab.ctx?'<div class="ctx">'+esc(tab.ctx)+'</div>':'')+
-      states+'</section>';
-    root.appendChild(el(sec));
-  });
-})();
-</script>
-</body>
-</html>
@@ -1,25 +0,0 @@
-// Render each .tab-section of a montage HTML into its own PNG (the PR-ready image).
-// Run from frontend/editor (so @playwright/test resolves):
-//   node <skill>/shoot-sections.mjs <montage.html> <outDir>
-import path from "node:path";
-import { pathToFileURL } from "node:url";
-import { createRequire } from "node:module";
-
-const require = createRequire(path.join(process.cwd(), "noop.js"));
-const { chromium } = require("@playwright/test");
-
-const htmlPath = path.resolve(process.argv[2]);
-const outDir = path.resolve(process.argv[3] || path.dirname(htmlPath));
-
-const browser = await chromium.launch();
-const page = await browser.newPage({ viewport: { width: 1200, height: 1200 }, deviceScaleFactor: 2 });
-await page.goto(pathToFileURL(htmlPath).href, { waitUntil: "load" });
-await page.waitForTimeout(250); // let images/fonts paint
-const ids = await page.$$eval(".tab-section", (els) => els.map((e) => e.id));
-if (!ids.length) { console.error("no .tab-section found"); process.exit(1); }
-for (const id of ids) {
-  const name = id.replace(/^section-/, "");
-  await page.locator("#" + id).screenshot({ path: path.join(outDir, `montage_${name}.png`) });
-  console.log("wrote montage_" + name + ".png");
-}
-await browser.close();
@@ -1,120 +0,0 @@
---
-name: ui-walkthrough
-description: >-
-  Full UI investigation of the current branch's feature. Enumerates every view
-  and state (empty, populated, loading, error, each dialog/menu/panel, responsive
-  breakpoints, light + dark + RTL), captures them with the stubbed Playwright
-  harness, assembles a single-image HTML walkthrough with a global light/dark
-  toggle slider, then runs two review passes: visual/consistency (alignment,
-  spacing, professionalism, dark/light parity, contrast, truncation) and
-  UX/ease-of-use (flow, discoverability, affordances, empty/error states,
-  expectations). Use when asked for a UI walkthrough, screenshot review, design
-  or QA pass, "find anywhere to make it easier/better for users", or before
-  merging frontend work. Pass --fix to auto-apply safe frontend fixes and
-  re-capture; --theme to limit themes; --no-rtl to skip RTL.
-argument-hint: "[feature/area] [--fix] [--theme light|dark|both] [--no-rtl] [--breakpoints]"
-allowed-tools: Read, Write, Edit, Glob, Grep, Bash
---
-
-# UI Walkthrough
-
-Produce a reviewable HTML walkthrough of a feature's UI in every state and theme,
-then critique it. Optionally auto-fix and re-capture.
-
-`$ARGUMENTS` may name the feature/area to focus on. If empty, scope from the
-current branch diff. Flags: `--fix`, `--theme light|dark|both` (default both),
-`--no-rtl`, `--breakpoints` (also capture phone/narrow widths).
-
-## What this repo gives you (use it, don't reinvent)
-
- **Stubbed Playwright project** = backend-free screenshots via `page.route()` mocks.
-  Reference implementation: `frontend/editor/src/core/tests/stubbed/files-page-screenshots.spec.ts`.
-  It already shows the light / **dark** / **RTL** passes, JWT seeding, IndexedDB
-  seeding, and dumping PNGs to a `screenshots/<area>/` folder. Copy its shape.
- Helpers: `frontend/editor/src/core/tests/helpers/ui-helpers.ts`
-  (`uploadFiles`, `openSettings`, `waitForModalOpen`, `dismissTourTooltip`, …)
-  and the `stub-test-base` fixtures (`autoGoto`, `seedJwt`, `viewport`).
- Config: `frontend/editor/playwright.config.ts` (run from `frontend/editor/`).
- Report template: [report-template.html](report-template.html) - self-contained,
-  one big image at a time, a global light/dark slider that flips every shot,
-  thumbnail rail, prev/next + arrow keys, and a Findings tab.
-
-## Process
-
-### 1. Scope the feature
- If `$ARGUMENTS` is empty: `git diff --name-only main...HEAD` and read the PR/commits.
-  Identify changed pages, tools (`core/components/tools/<tool>` or `core/tools/<tool>`),
-  dialogs, panels, and routes.
- Enumerate **every view and state** to capture, e.g.:
-  empty / populated / loading / error / disabled; each dialog, menu, popover, tooltip;
-  each tab or step; selection + multi-select; success/result panel; and (if relevant)
-  permission/role variants. Write the list down before capturing - it's the report's spine.
-
-### 2. Prepare the harness (worktree-safe)
-Worktrees have no `node_modules` and no generated icons. From repo root:
-```
-cd frontend && npm ci                       # or junction main's node_modules (see memory)
-cd frontend/editor && node scripts/generate-icons.js
-```
-Kill any stale dev server first (it serves old modules):
-`Get-NetTCPConnection -LocalPort 5173 -State Listen | %{ Stop-Process -Id $_.OwningProcess -Force }`
-
-### 3. Write the capture spec
-Create `frontend/editor/src/core/tests/stubbed/<feature>-walkthrough.spec.ts`,
-modeled on `files-page-screenshots.spec.ts`. For each enumerated view:
- stub the APIs it needs, drive the UI to that state, wait on a real locator
-  (not a fixed sleep), `await settle(page)` for Mantine portals, then
-  `page.screenshot({ path: shotPath("NN_name_<theme>") })`.
- Capture each view in **light and dark** (and RTL unless `--no-rtl`). Reuse the
-  `enableDarkMode` / `enableRtl` init-script pattern from the reference spec
-  (`localStorage["mantine-color-scheme"]="dark"` + `emulateMedia({colorScheme:"dark"})`).
- Name shots `NN_<view>_<theme>.png` so light/dark pair up by suffix.
- Prefer **stable test-ids** over translated accessible names (RTL/i18n breaks text locators).
-
-Run it: `cd frontend/editor && npx playwright test --project=stubbed <feature>-walkthrough.spec.ts`.
-Add `--project=stubbed-firefox`/`-webkit` only if cross-browser layout matters.
-
-### 4. Build the report
- Copy `report-template.html` to `screenshots/<feature>/walkthrough.html` (so the
-  relative `screenshots/...` image paths resolve, or rewrite paths to sit beside it).
- Build the manifest and inject it: replace the JSON between the
-  `/*__DATA__*/` … `/*__END__*/` markers with one `views[]` entry per view
-  (`{id,title,light,dark,viewport,notes}`) and an empty `findings` object you'll
-  fill in step 5. Keep `light`/`dark` as relative paths.
- The toggle slider answers the "one big image + flip light/dark for all" request:
-  it shows a single large screenshot, and switching the slider re-themes every view.
-
-### 5. Review pass 1 - visual & consistency
-Open each screenshot (Read the PNG) and judge against the others:
-alignment & spacing rhythm, control placement, button hierarchy, typography,
-**light/dark parity** (contrast, invisible borders, washed-out text, wrong tokens),
-truncation/overflow, RTL mirroring, focus states, icon consistency, professional polish.
-Record each issue as a finding `{severity:high|med|low, view, title, detail, fix}`.
-
-### 6. Review pass 2 - UX & ease of use
-Walk the flow as a first-time user: discoverability, number of steps, affordance
-clarity, empty-state guidance, error recovery, destructive-action confirmation,
-defaults, loading feedback, mobile reachability, accessible names, and whether the
-UI matches user expectations for this kind of tool. Record findings the same way.
-
-Write both finding lists into the report's `findings.visual` / `findings.ux`,
-and add short per-view `notes`. Re-inject the manifest.
-
-### 7. If `--fix`
-Only safe, self-contained frontend fixes (spacing, alignment, tokens, missing
-dark-mode colors, labels, aria, obvious copy). For each: edit the component/CSS,
-mark the finding `fixed:true` with what changed, then **re-run the spec** to
-re-capture the affected shots and regenerate the report. Run `task frontend:check`.
-Leave anything risky or ambiguous as a finding, not a change.
-
-### 8. Deliver
-Tell the user the report path and give a tight chat summary: N views ×
-themes captured, top findings by severity, and (if `--fix`) what changed.
-Optionally `SendUserFile` the `walkthrough.html`.
-
-## Gotchas
- Stale `:5173` server serves old bundles - kill it before capturing (see step 2).
- Missing `material-symbols-icons.json` → blank app → every shot times out. Run
-  `generate-icons.js` first.
- `await settle(page)` before shots or portals/transitions tear mid-capture.
- Don't commit the generated `screenshots/` or the throwaway spec unless asked.
@@ -1,116 +0,0 @@
-"""Build a self-contained EXAMPLE.html from report-template.html with mock
-light/dark screenshots, so the viewer + global theme slider can be demoed
-without a real capture run. Run: python make_example.py"""
-import base64
-import json
-import pathlib
-import re
-
-HERE = pathlib.Path(__file__).parent
-
-
-def svg(bg, fg, panel, accent, muted, label, kind):
-    """A simple fake 'screen' SVG: title bar, sidebar, content varies by kind."""
-    parts = [
-        f'<svg xmlns="http://www.w3.org/2000/svg" width="1600" height="900" viewBox="0 0 1600 900">',
-        f'<rect width="1600" height="900" fill="{bg}"/>',
-        # top bar
-        f'<rect width="1600" height="64" fill="{panel}"/>',
-        f'<circle cx="40" cy="32" r="12" fill="{accent}"/>',
-        f'<rect x="64" y="24" width="160" height="16" rx="6" fill="{muted}"/>',
-        f'<rect x="1430" y="20" width="130" height="24" rx="12" fill="{accent}"/>',
-        # left sidebar
-        f'<rect x="0" y="64" width="220" height="836" fill="{panel}"/>',
-    ]
-    for i in range(6):
-        y = 100 + i * 56
-        parts.append(f'<rect x="24" y="{y}" width="172" height="32" rx="8" fill="{bg}"/>')
-    if kind == "empty":
-        parts += [
-            f'<rect x="700" y="360" width="200" height="120" rx="16" fill="none" stroke="{muted}" stroke-width="3" stroke-dasharray="10 8"/>',
-            f'<rect x="690" y="510" width="220" height="44" rx="10" fill="{accent}"/>',
-            f'<text x="800" y="600" fill="{muted}" font-family="sans-serif" font-size="26" text-anchor="middle">{label}</text>',
-        ]
-    elif kind == "form":
-        for i in range(4):
-            y = 140 + i * 90
-            parts.append(f'<rect x="280" y="{y}" width="160" height="16" rx="6" fill="{muted}"/>')
-            parts.append(f'<rect x="280" y="{y+26}" width="900" height="44" rx="8" fill="{panel}" stroke="{muted}" stroke-width="1"/>')
-        parts.append(f'<rect x="280" y="560" width="200" height="50" rx="10" fill="{accent}"/>')
-        parts.append(f'<text x="800" y="850" fill="{muted}" font-family="sans-serif" font-size="24" text-anchor="middle">{label}</text>')
-    else:  # dialog
-        parts += [
-            f'<rect width="1600" height="900" fill="{fg}" opacity="0.45"/>',
-            f'<rect x="520" y="280" width="560" height="360" rx="18" fill="{panel}"/>',
-            f'<rect x="556" y="320" width="280" height="22" rx="8" fill="{fg}"/>',
-            f'<rect x="556" y="372" width="488" height="14" rx="6" fill="{muted}"/>',
-            f'<rect x="556" y="398" width="420" height="14" rx="6" fill="{muted}"/>',
-            f'<rect x="820" y="560" width="110" height="44" rx="9" fill="{bg}" stroke="{muted}"/>',
-            f'<rect x="946" y="560" width="98" height="44" rx="9" fill="{accent}"/>',
-            f'<text x="800" y="700" fill="#fff" font-family="sans-serif" font-size="24" text-anchor="middle">{label}</text>',
-        ]
-    parts.append("</svg>")
-    return "".join(parts)
-
-
-def data_uri(s):
-    return "data:image/svg+xml;base64," + base64.b64encode(s.encode()).decode()
-
-
-LIGHT = dict(bg="#ffffff", fg="#111418", panel="#f1f3f6", accent="#2f6fed", muted="#c2c8d0")
-DARK = dict(bg="#16181c", fg="#000000", panel="#1f232a", accent="#5b8cff", muted="#3a414b")
-
-
-def pair(kind, label):
-    return (
-        data_uri(svg(LIGHT["bg"], LIGHT["fg"], LIGHT["panel"], LIGHT["accent"], LIGHT["muted"], label, kind)),
-        data_uri(svg(DARK["bg"], DARK["fg"], DARK["panel"], DARK["accent"], DARK["muted"], label, kind)),
-    )
-
-
-views = []
-for idx, (kind, title, label) in enumerate([
-    ("empty", "Empty state", "Drop a PDF to start"),
-    ("form", "Tool options panel", "Compress options"),
-    ("dialog", "Confirm dialog", "Replace original file?"),
-], start=1):
-    light, dark = pair(kind, label)
-    views.append({
-        "id": f"{idx:02d}_{kind}",
-        "title": title,
-        "light": light,
-        "dark": dark,
-        "viewport": "1600x900",
-        "notes": ["This is mock data to demo the viewer."],
-    })
-
-data = {
-    "feature": "EXAMPLE - Compress PDF (mock data)",
-    "branch": "demo",
-    "generated": "example",
-    "views": views,
-    "findings": {
-        "visual": [
-            {"severity": "high", "view": "03_dialog", "title": "Dialog buttons too close",
-             "detail": "Cancel/Confirm have only 8px gap; easy to misclick.",
-             "fix": "Increase gap to var(--mantine-spacing-md)."},
-            {"severity": "low", "view": "02_form", "title": "Field labels low contrast in dark mode",
-             "detail": "Muted token fails WCAG AA on the dark panel.",
-             "fix": "Use --mantine-color-dimmed instead of a hard-coded grey."},
-        ],
-        "ux": [
-            {"severity": "med", "view": "01_empty", "title": "Primary CTA below the dropzone",
-             "detail": "Users expect the action button adjacent to the dropzone.",
-             "fix": "Move the button directly under the dashed zone."},
-        ],
-    },
-}
-
-tpl = (HERE / "report-template.html").read_text(encoding="utf-8")
-out = re.sub(
-    r"/\*__DATA__\*/.*?/\*__END__\*/",
-    lambda _m: "/*__DATA__*/" + json.dumps(data) + "/*__END__*/",
-    tpl, count=1, flags=re.S,
-)
-(HERE / "EXAMPLE.html").write_text(out, encoding="utf-8")
-print("wrote", (HERE / "EXAMPLE.html"))
@@ -1,298 +0,0 @@
-<!doctype html>
-<!--
-  UI Walkthrough report template (self-contained, works from file://).
-  The ui-walkthrough skill replaces the JSON in the window.__WALKTHROUGH__ data
-  block below with the captured manifest. Do not add external CDN deps - it must open offline.
-
-  Data shape:
-  {
-    "feature": "Compress PDF tool",
-    "branch": "claude/...",
-    "generated": "2026-06-21",
-    "views": [
-      { "id": "01_empty", "title": "Empty state",
-        "light": "screenshots/compress/01_empty_light.png",
-        "dark":  "screenshots/compress/01_empty_dark.png",
-        "viewport": "1600x900",
-        "notes": ["Heading is centered", "Primary CTA below the fold on mobile"] }
-    ],
-    "findings": {
-      "visual": [ { "severity":"high", "view":"01_empty", "title":"...", "detail":"...", "fix":"..." } ],
-      "ux":     [ { "severity":"med",  "view":"03_dialog", "title":"...", "detail":"...", "fix":"..." } ]
-    }
-  }
-->
-<html lang="en">
-<head>
-<meta charset="utf-8" />
-<meta name="viewport" content="width=device-width, initial-scale=1" />
-<title>UI Walkthrough</title>
-<style>
-  :root {
-    --bg: #f6f7f9; --panel: #ffffff; --panel-2: #f0f2f5; --text: #1a1b1e;
-    --muted: #6b7280; --border: #e2e5ea; --accent: #2f6fed; --accent-weak: #e8f0fe;
-    --shadow: 0 1px 3px rgba(0,0,0,.08), 0 8px 24px rgba(0,0,0,.06);
-    --hi: #d92d20; --med: #d98e00; --low: #2f6fed; --stage: #0b0c0e;
-  }
-  html[data-theme="dark"] {
-    --bg: #0d0e10; --panel: #16181c; --panel-2: #1d2024; --text: #e6e8eb;
-    --muted: #9aa3ad; --border: #2a2e35; --accent: #5b8cff; --accent-weak: #1a2336;
-    --shadow: 0 1px 3px rgba(0,0,0,.5), 0 8px 24px rgba(0,0,0,.4); --stage: #000;
-  }
-  * { box-sizing: border-box; }
-  body { margin: 0; font: 14px/1.5 -apple-system, "Segoe UI", Roboto, system-ui, sans-serif;
-         background: var(--bg); color: var(--text); }
-  header { display: flex; align-items: center; gap: 16px; padding: 12px 20px;
-           background: var(--panel); border-bottom: 1px solid var(--border); position: sticky; top: 0; z-index: 5; }
-  header h1 { font-size: 15px; margin: 0; font-weight: 650; }
-  header .sub { color: var(--muted); font-size: 12px; }
-  .spacer { flex: 1; }
-  .counter { color: var(--muted); font-variant-numeric: tabular-nums; font-size: 13px; }
-  .tabs { display: flex; gap: 4px; }
-  .tab { border: 1px solid var(--border); background: var(--panel-2); color: var(--text);
-         padding: 6px 12px; border-radius: 8px; cursor: pointer; font-size: 13px; }
-  .tab.active { background: var(--accent); color: #fff; border-color: var(--accent); }
-
-  /* Light/Dark slider */
-  .theme-toggle { display: flex; align-items: center; gap: 9px; user-select: none; }
-  .theme-toggle .lbl { font-size: 12px; color: var(--muted); }
-  .theme-toggle .lbl.on { color: var(--text); font-weight: 600; }
-  .switch { position: relative; width: 52px; height: 28px; }
-  .switch input { opacity: 0; width: 0; height: 0; }
-  .slider { position: absolute; inset: 0; cursor: pointer; background: var(--panel-2);
-            border: 1px solid var(--border); border-radius: 999px; transition: .2s; }
-  .slider:before { content: ""; position: absolute; height: 20px; width: 20px; left: 3px; top: 3px;
-                   background: #fbbf24; border-radius: 50%; transition: .2s; box-shadow: 0 1px 2px rgba(0,0,0,.3); }
-  .switch input:checked + .slider { background: var(--accent); }
-  .switch input:checked + .slider:before { transform: translateX(24px); background: #c7d2fe; }
-
-  main { display: grid; grid-template-columns: 240px 1fr; height: calc(100vh - 53px); }
-  .rail { border-right: 1px solid var(--border); overflow-y: auto; background: var(--panel); padding: 8px; }
-  .rail .group-label { font-size: 11px; text-transform: uppercase; letter-spacing: .05em;
-                       color: var(--muted); padding: 10px 8px 4px; }
-  .thumb { display: flex; gap: 9px; align-items: center; padding: 7px; border-radius: 8px;
-           cursor: pointer; border: 1px solid transparent; }
-  .thumb:hover { background: var(--panel-2); }
-  .thumb.active { background: var(--accent-weak); border-color: var(--accent); }
-  .thumb img { width: 64px; height: 40px; object-fit: cover; border-radius: 4px; border: 1px solid var(--border); background: var(--stage); }
-  .thumb .t { font-size: 12.5px; line-height: 1.3; }
-  .thumb .badge { font-size: 10px; color: var(--muted); }
-  .thumb .dot { width: 7px; height: 7px; border-radius: 50%; margin-left: auto; flex: none; }
-
-  .stagewrap { display: flex; flex-direction: column; min-width: 0; }
-  .stage { flex: 1; display: flex; align-items: center; justify-content: center; padding: 22px;
-           background: var(--stage); position: relative; min-height: 0; }
-  .stage img { max-width: 100%; max-height: 100%; object-fit: contain; border-radius: 8px;
-               box-shadow: 0 4px 30px rgba(0,0,0,.4); background: #fff; }
-  html[data-theme="dark"] .stage img { background: #16181c; }
-  .nav-btn { position: absolute; top: 50%; transform: translateY(-50%); width: 42px; height: 42px;
-             border-radius: 50%; border: 1px solid var(--border); background: var(--panel);
-             color: var(--text); cursor: pointer; font-size: 18px; opacity: .85; }
-  .nav-btn:hover { opacity: 1; } .nav-btn.prev { left: 16px; } .nav-btn.next { right: 16px; }
-  .nav-btn:disabled { opacity: .25; cursor: default; }
-  .missing { color: var(--muted); font-size: 13px; text-align: center; }
-
-  .detail { border-top: 1px solid var(--border); background: var(--panel); padding: 14px 20px;
-            max-height: 38vh; overflow-y: auto; }
-  .detail h2 { margin: 0 0 4px; font-size: 15px; }
-  .detail .meta { color: var(--muted); font-size: 12px; margin-bottom: 10px; }
-  .notes { list-style: none; padding: 0; margin: 0; display: grid; gap: 6px; }
-  .notes li { display: flex; gap: 8px; align-items: flex-start; }
-  .sev { font-size: 10px; font-weight: 700; text-transform: uppercase; padding: 2px 7px; border-radius: 999px;
-         color: #fff; flex: none; margin-top: 1px; }
-  .sev.high { background: var(--hi); } .sev.med { background: var(--med); } .sev.low { background: var(--low); }
-  .finding .fix { color: var(--muted); font-size: 12.5px; }
-  .finding .fix b { color: var(--text); font-weight: 600; }
-
-  /* Summary tab */
-  .summary { padding: 20px 28px; overflow-y: auto; }
-  .summary h2 { font-size: 16px; margin: 22px 0 8px; }
-  .summary .empty { color: var(--muted); }
-  .card { background: var(--panel); border: 1px solid var(--border); border-radius: 10px;
-          padding: 12px 14px; margin-bottom: 8px; box-shadow: var(--shadow); }
-  .card .head { display: flex; gap: 8px; align-items: center; }
-  .card a { color: var(--accent); text-decoration: none; cursor: pointer; }
-  .hide { display: none !important; }
-  kbd { font: 11px ui-monospace, monospace; background: var(--panel-2); border: 1px solid var(--border);
-        border-radius: 4px; padding: 1px 5px; }
-</style>
-</head>
-<body>
-<header>
-  <div>
-    <h1 id="feature-title">UI Walkthrough</h1>
-    <div class="sub" id="feature-sub"></div>
-  </div>
-  <div class="spacer"></div>
-  <div class="tabs">
-    <button class="tab active" data-tab="viewer">Walkthrough</button>
-    <button class="tab" data-tab="summary">Findings</button>
-  </div>
-  <div class="counter" id="counter"></div>
-  <label class="theme-toggle" title="Toggle light / dark for every screenshot">
-    <span class="lbl" id="lbl-light">Light</span>
-    <span class="switch"><input type="checkbox" id="theme-switch" /><span class="slider"></span></span>
-    <span class="lbl" id="lbl-dark">Dark</span>
-  </label>
-</header>
-
-<main id="viewer-pane">
-  <aside class="rail" id="rail"></aside>
-  <section class="stagewrap">
-    <div class="stage">
-      <button class="nav-btn prev" id="prev" aria-label="Previous">&#8249;</button>
-      <img id="stage-img" alt="" />
-      <div class="missing hide" id="missing"></div>
-      <button class="nav-btn next" id="next" aria-label="Next">&#8250;</button>
-    </div>
-    <div class="detail">
-      <h2 id="view-title"></h2>
-      <div class="meta" id="view-meta"></div>
-      <ul class="notes" id="view-notes"></ul>
-    </div>
-  </section>
-</main>
-
-<section class="summary hide" id="summary-pane"></section>
-
-<script id="data">
-window.__WALKTHROUGH__ = /*__DATA__*/{"feature":"No data","branch":"","generated":"","views":[],"findings":{"visual":[],"ux":[]}}/*__END__*/;
-</script>
-<script>
-(function () {
-  var D = window.__WALKTHROUGH__ || { views: [], findings: { visual: [], ux: [] } };
-  var views = D.views || [];
-  var state = { i: 0, theme: localStorage.getItem("ui-wt-theme") || "light", tab: "viewer" };
-
-  var $ = function (id) { return document.getElementById(id); };
-  function sevClass(s) { return s === "high" ? "high" : s === "med" || s === "medium" ? "med" : "low"; }
-
-  function applyChrome() {
-    document.documentElement.setAttribute("data-theme", state.theme);
-    $("theme-switch").checked = state.theme === "dark";
-    $("lbl-light").classList.toggle("on", state.theme === "light");
-    $("lbl-dark").classList.toggle("on", state.theme === "dark");
-  }
-
-  function srcFor(v) { return state.theme === "dark" ? (v.dark || v.light) : (v.light || v.dark); }
-
-  function findingsForView(id) {
-    var all = (D.findings && D.findings.visual || []).concat(D.findings && D.findings.ux || []);
-    return all.filter(function (f) { return f.view === id; });
-  }
-
-  function renderRail() {
-    var rail = $("rail");
-    rail.innerHTML = "";
-    if (!views.length) { rail.innerHTML = '<div class="group-label">No views captured</div>'; return; }
-    views.forEach(function (v, idx) {
-      var fs = findingsForView(v.id);
-      var worst = fs.some(function (f){return sevClass(f.severity)==="high";}) ? "var(--hi)"
-                : fs.some(function (f){return sevClass(f.severity)==="med";}) ? "var(--med)"
-                : fs.length ? "var(--low)" : "transparent";
-      var el = document.createElement("div");
-      el.className = "thumb" + (idx === state.i ? " active" : "");
-      el.innerHTML = '<img src="' + srcFor(v) + '" alt="" />' +
-        '<div><div class="t">' + (v.title || v.id) + '</div>' +
-        '<div class="badge">' + (v.viewport || "") + '</div></div>' +
-        '<span class="dot" style="background:' + worst + '"></span>';
-      el.onclick = function () { state.i = idx; render(); };
-      rail.appendChild(el);
-    });
-  }
-
-  function render() {
-    applyChrome();
-    if (!views.length) {
-      $("missing").classList.remove("hide"); $("stage-img").classList.add("hide");
-      $("missing").textContent = "No screenshots in this report yet.";
-      $("counter").textContent = ""; return;
-    }
-    var v = views[state.i];
-    var src = srcFor(v);
-    var img = $("stage-img");
-    if (src) {
-      img.classList.remove("hide"); $("missing").classList.add("hide");
-      img.src = src; img.alt = v.title || v.id;
-    } else {
-      img.classList.add("hide"); $("missing").classList.remove("hide");
-      $("missing").textContent = "No " + state.theme + " screenshot for this view.";
-    }
-    $("counter").textContent = (state.i + 1) + " / " + views.length;
-    $("view-title").textContent = v.title || v.id;
-    $("view-meta").textContent = [v.viewport, state.theme + " mode"].filter(Boolean).join("  ·  ");
-    var notes = $("view-notes"); notes.innerHTML = "";
-    var fs = findingsForView(v.id);
-    (v.notes || []).forEach(function (n) {
-      var li = document.createElement("li"); li.textContent = "· " + n; notes.appendChild(li);
-    });
-    fs.forEach(function (f) {
-      var li = document.createElement("li"); li.className = "finding";
-      li.innerHTML = '<span class="sev ' + sevClass(f.severity) + '">' + (f.severity || "note") + '</span>' +
-        '<span><b>' + (f.title || "") + '</b> — ' + (f.detail || "") +
-        (f.fix ? ' <span class="fix"><b>Fix:</b> ' + f.fix + '</span>' : '') + '</span>';
-      notes.appendChild(li);
-    });
-    $("prev").disabled = state.i === 0;
-    $("next").disabled = state.i === views.length - 1;
-    renderRail();
-  }
-
-  function renderSummary() {
-    var pane = $("summary-pane");
-    function block(title, arr) {
-      var h = '<h2>' + title + ' (' + arr.length + ')</h2>';
-      if (!arr.length) return h + '<div class="empty">None found.</div>';
-      return h + arr.map(function (f) {
-        return '<div class="card"><div class="head">' +
-          '<span class="sev ' + sevClass(f.severity) + '">' + (f.severity || "note") + '</span>' +
-          '<b>' + (f.title || "") + '</b>' +
-          (f.view ? ' <a data-jump="' + f.view + '">' + f.view + '</a>' : '') + '</div>' +
-          '<div style="margin-top:6px">' + (f.detail || "") + '</div>' +
-          (f.fix ? '<div class="finding" style="margin-top:6px"><span class="fix"><b>Fix:</b> ' + f.fix + '</span></div>' : '') +
-          '</div>';
-      }).join("");
-    }
-    pane.innerHTML = block("Visual & consistency", (D.findings && D.findings.visual) || []) +
-                     block("UX & ease of use", (D.findings && D.findings.ux) || []);
-    pane.querySelectorAll("[data-jump]").forEach(function (a) {
-      a.onclick = function () {
-        var id = a.getAttribute("data-jump");
-        var idx = views.findIndex(function (v) { return v.id === id; });
-        if (idx >= 0) { state.i = idx; setTab("viewer"); }
-      };
-    });
-  }
-
-  function setTab(t) {
-    state.tab = t;
-    document.querySelectorAll(".tab").forEach(function (b) { b.classList.toggle("active", b.dataset.tab === t); });
-    $("viewer-pane").classList.toggle("hide", t !== "viewer");
-    $("summary-pane").classList.toggle("hide", t !== "summary");
-    if (t === "viewer") $("viewer-pane").style.display = "grid";
-    if (t === "summary") renderSummary();
-  }
-
-  // wiring
-  $("feature-title").textContent = D.feature || "UI Walkthrough";
-  $("feature-sub").textContent = [D.branch, D.generated].filter(Boolean).join("  ·  ");
-  $("theme-switch").onchange = function () {
-    state.theme = this.checked ? "dark" : "light";
-    localStorage.setItem("ui-wt-theme", state.theme);
-    render();
-  };
-  $("prev").onclick = function () { if (state.i > 0) { state.i--; render(); } };
-  $("next").onclick = function () { if (state.i < views.length - 1) { state.i++; render(); } };
-  document.addEventListener("keydown", function (e) {
-    if (state.tab !== "viewer") return;
-    if (e.key === "ArrowLeft") $("prev").click();
-    if (e.key === "ArrowRight") $("next").click();
-    if (e.key.toLowerCase() === "t") $("theme-switch").click();
-  });
-  document.querySelectorAll(".tab").forEach(function (b) { b.onclick = function () { setTab(b.dataset.tab); }; });
-
-  render();
-})();
-</script>
-</body>
-</html>
@@ -20,8 +20,8 @@ set -e
 # - To build the project, use:
 #     ./gradlew build
 #
-# - To run the lint/format/secret checks, use:
-#     task pre-commit
+# - For running pre-commit hooks (if configured), use:
+#     pre-commit run --all-files
 #
 # Make sure you are in the project root directory after this script executes.
 # =============================================================================
@@ -70,6 +70,6 @@ echo ""
 echo "  To build the project: "
 echo -e "\e[34m    gradle build\e[0m"
 echo ""
-echo "  To run the lint/format/secret checks:"
-echo -e "\e[34m    task pre-commit\e[0m"
+echo "  To run pre-commit hooks (if configured):"
+echo -e "\e[34m    pre-commit run --all-files -c .pre-commit-config.yaml\e[0m"
 echo "=================================================================="
@@ -27,6 +27,7 @@ node_modules/
 **/node_modules/
 frontend/node_modules/
 frontend/editor/dist/
+frontend/dist-portal/
 frontend/editor/playwright-report/
 .npm/
 .yarn/
@@ -1,6 +1,6 @@
 # Maintainer: Stirling PDF Inc <contact@stirlingpdf.com>
 pkgname=stirling-pdf-desktop
-pkgver=2.14.1
+pkgver=2.12.0
 pkgrel=1
 pkgdesc="Locally hosted, web-based PDF manipulation tool (Tauri desktop app, official Stirling PDF Inc build)"
 arch=('x86_64')
@@ -1,6 +1,6 @@
 # Maintainer: Stirling PDF Inc <contact@stirlingpdf.com>
 pkgname=stirling-pdf-server-bin
-pkgver=2.14.1
+pkgver=2.12.0
 pkgrel=1
 pkgdesc="Locally hosted, web-based PDF manipulation tool (server JAR, prebuilt)"
 arch=('any')
@@ -1,8 +1,6 @@
 build: &build
  - build.gradle
  - app/(common|core|proprietary)/build.gradle
-  - Taskfile.yml
-  - .taskfiles/backend.yml

 openapi: &openapi
  - *build
@@ -40,9 +38,6 @@ project: &project
  - frontend/**
  - docker/**
  - scripts/RestartHelper.java
-  - Taskfile.yml
-  - .taskfiles/backend.yml
-  - .taskfiles/docker.yml
  - scripts/db-migration/**
  - .github/workflows/db-migration-test.yml

@@ -60,9 +55,6 @@ frontend: &frontend
  - scripts/summarize_type3_signatures.py
  - scripts/type3_to_cff.py
  - scripts/update_type3_library.py
-  - Taskfile.yml
-  - .taskfiles/frontend.yml
-  - .taskfiles/e2e.yml

 # Files that affect the Tauri desktop bundle. Gate the multi-OS Tauri build
 # job on changes to any of these.
@@ -74,8 +66,6 @@ tauri: &tauri
  - frontend/package-lock.json
  - frontend/editor/vite.config.ts
  - .github/workflows/tauri-build.yml
-  - Taskfile.yml
-  - .taskfiles/desktop.yml

 # Files that affect the AI engine (Python tool models, fixers, tests). Gate
 # the engine validation job on changes to engine sources or to the Java
@@ -84,23 +74,6 @@ engine: &engine
  - engine/**
  - app/(common|core|proprietary)/src/main/java/**
  - .github/workflows/ai-engine.yml
-  - Taskfile.yml
-  - .taskfiles/engine.yml
-
-# Files that can make the committed generated API models (frontend tool API
-# types + engine tool models) go stale: the Java tool surfaces they derive from,
-# the generators, the generated files themselves (to catch a hand-edit), and the
-# tasks that drive generation. Deliberately excludes the broad frontend/docker/
-# testing globs, so a CSS-only PR does not boot the backend to rebuild the spec.
-generated-models: &generated-models
-  - *openapi
-  - frontend/editor/scripts/generate-tool-api-types.mts
-  - frontend/editor/src/core/types/toolApiTypes.ts
-  - engine/scripts/generate_tool_models.py
-  - engine/src/stirling/models/tool_models.py
-  - .taskfiles/frontend.yml
-  - .taskfiles/engine.yml
-  - .github/workflows/check-generated-models.yml

 licenses-frontend: &licenses-frontend
  - ".github/workflows/frontend-backend-licenses-update.yml"
@@ -129,4 +102,4 @@ proprietary: &proprietary
  - configs/settings.yml.template
  - build.gradle
  - app/proprietary/build.gradle
-  - .github/workflows/build-enterprise.yml
+  - .github/workflows/build-enterprise.yml
@@ -13,7 +13,7 @@ Usage:
 """

 # Sample for Windows:
-# python .github/scripts/check_language_toml.py --reference-file frontend/editor/public/locales/en-US/translation.toml --branch "" --files frontend/editor/public/locales/de-DE/translation.toml frontend/editor/public/locales/fr-FR/translation.toml
+# python .github/scripts/check_language_toml.py --reference-file frontend/editor/public/locales/en-GB/translation.toml --branch "" --files frontend/editor/public/locales/de-DE/translation.toml frontend/editor/public/locales/fr-FR/translation.toml

 import argparse
 import glob
@@ -211,7 +211,7 @@ def check_for_differences(reference_file, file_list, branch, actor):
            )
            continue

-        if basename_current_file == basename_reference_file and locale_dir == "en-US":
+        if basename_current_file == basename_reference_file and locale_dir == "en-GB":
            continue

        if (
@@ -308,7 +308,7 @@ def check_for_differences(reference_file, file_list, branch, actor):
        report.append("## ❌ Overall Check Status: **_Failed_**")
        report.append("")
        report.append(
-            f"@{actor} please check your translation if it conforms to the standard. Follow the format of [en-US/translation.toml](https://github.com/Stirling-Tools/Stirling-PDF/blob/main/frontend/editor/public/locales/en-US/translation.toml)"
+            f"@{actor} please check your translation if it conforms to the standard. Follow the format of [en-GB/translation.toml](https://github.com/Stirling-Tools/Stirling-PDF/blob/main/frontend/editor/public/locales/en-GB/translation.toml)"
        )
    else:
        report.append("## ✅ Overall Check Status: **_Success_**")
@@ -0,0 +1 @@
+pre-commit
@@ -0,0 +1,121 @@
+#
+# This file is autogenerated by pip-compile with Python 3.12
+# by the following command:
+#
+#    pip-compile --generate-hashes --output-file='.github\scripts\requirements_pre_commit.txt' --strip-extras '.github\scripts\requirements_pre_commit.in'
+#
+cfgv==3.5.0 \
+    --hash=sha256:a8dc6b26ad22ff227d2634a65cb388215ce6cc96bbcc5cfde7641ae87e8dacc0 \
+    --hash=sha256:d5b1034354820651caa73ede66a6294d6e95c1b00acc5e9b098e917404669132
+    # via pre-commit
+distlib==0.4.0 \
+    --hash=sha256:9659f7d87e46584a30b5780e43ac7a2143098441670ff0a49d5f9034c54a6c16 \
+    --hash=sha256:feec40075be03a04501a973d81f633735b4b69f98b05450592310c0f401a4e0d
+    # via virtualenv
+filelock==3.29.0 \
+    --hash=sha256:69974355e960702e789734cb4871f884ea6fe50bd8404051a3530bc07809cf90 \
+    --hash=sha256:96f5f6344709aa1572bbf631c640e4ebeeb519e08da902c39a001882f30ac258
+    # via
+    #   python-discovery
+    #   virtualenv
+identify==2.6.19 \
+    --hash=sha256:20e6a87f786f768c092a721ad107fc9df0eb89347be9396cadf3f4abbd1fb78a \
+    --hash=sha256:6be5020c38fcb07da56c53733538a3081ea5aa70d36a156f83044bfbf9173842
+    # via pre-commit
+nodeenv==1.10.0 \
+    --hash=sha256:5bb13e3eed2923615535339b3c620e76779af4cb4c6a90deccc9e36b274d3827 \
+    --hash=sha256:996c191ad80897d076bdfba80a41994c2b47c68e224c542b48feba42ba00f8bb
+    # via pre-commit
+platformdirs==4.9.6 \
+    --hash=sha256:3bfa75b0ad0db84096ae777218481852c0ebc6c727b3168c1b9e0118e458cf0a \
+    --hash=sha256:e61adb1d5e5cb3441b4b7710bea7e4c12250ca49439228cc1021c00dcfac0917
+    # via
+    #   python-discovery
+    #   virtualenv
+pre-commit==4.6.0 \
+    --hash=sha256:718d2208cef53fdc38206e40524a6d4d9576d103eb16f0fec11c875e7716e9d9 \
+    --hash=sha256:e2cf246f7299edcabcf15f9b0571fdce06058527f0a06535068a86d38089f29b
+    # via -r .github/scripts/requirements_pre_commit.in
+python-discovery==1.2.2 \
+    --hash=sha256:876e9c57139eb757cb5878cbdd9ae5379e5d96266c99ef731119e04fffe533bb \
+    --hash=sha256:e1ae95d9af875e78f15e19aed0c6137ab1bb49c200f21f5061786490c9585c7a
+    # via virtualenv
+pyyaml==6.0.3 \
+    --hash=sha256:00c4bdeba853cc34e7dd471f16b4114f4162dc03e6b7afcc2128711f0eca823c \
+    --hash=sha256:0150219816b6a1fa26fb4699fb7daa9caf09eb1999f3b70fb6e786805e80375a \
+    --hash=sha256:02893d100e99e03eda1c8fd5c441d8c60103fd175728e23e431db1b589cf5ab3 \
+    --hash=sha256:02ea2dfa234451bbb8772601d7b8e426c2bfa197136796224e50e35a78777956 \
+    --hash=sha256:0f29edc409a6392443abf94b9cf89ce99889a1dd5376d94316ae5145dfedd5d6 \
+    --hash=sha256:10892704fc220243f5305762e276552a0395f7beb4dbf9b14ec8fd43b57f126c \
+    --hash=sha256:16249ee61e95f858e83976573de0f5b2893b3677ba71c9dd36b9cf8be9ac6d65 \
+    --hash=sha256:1d37d57ad971609cf3c53ba6a7e365e40660e3be0e5175fa9f2365a379d6095a \
+    --hash=sha256:1ebe39cb5fc479422b83de611d14e2c0d3bb2a18bbcb01f229ab3cfbd8fee7a0 \
+    --hash=sha256:214ed4befebe12df36bcc8bc2b64b396ca31be9304b8f59e25c11cf94a4c033b \
+    --hash=sha256:2283a07e2c21a2aa78d9c4442724ec1eb15f5e42a723b99cb3d822d48f5f7ad1 \
+    --hash=sha256:22ba7cfcad58ef3ecddc7ed1db3409af68d023b7f940da23c6c2a1890976eda6 \
+    --hash=sha256:27c0abcb4a5dac13684a37f76e701e054692a9b2d3064b70f5e4eb54810553d7 \
+    --hash=sha256:28c8d926f98f432f88adc23edf2e6d4921ac26fb084b028c733d01868d19007e \
+    --hash=sha256:2e71d11abed7344e42a8849600193d15b6def118602c4c176f748e4583246007 \
+    --hash=sha256:34d5fcd24b8445fadc33f9cf348c1047101756fd760b4dacb5c3e99755703310 \
+    --hash=sha256:37503bfbfc9d2c40b344d06b2199cf0e96e97957ab1c1b546fd4f87e53e5d3e4 \
+    --hash=sha256:3c5677e12444c15717b902a5798264fa7909e41153cdf9ef7ad571b704a63dd9 \
+    --hash=sha256:3ff07ec89bae51176c0549bc4c63aa6202991da2d9a6129d7aef7f1407d3f295 \
+    --hash=sha256:41715c910c881bc081f1e8872880d3c650acf13dfa8214bad49ed4cede7c34ea \
+    --hash=sha256:418cf3f2111bc80e0933b2cd8cd04f286338bb88bdc7bc8e6dd775ebde60b5e0 \
+    --hash=sha256:44edc647873928551a01e7a563d7452ccdebee747728c1080d881d68af7b997e \
+    --hash=sha256:4a2e8cebe2ff6ab7d1050ecd59c25d4c8bd7e6f400f5f82b96557ac0abafd0ac \
+    --hash=sha256:4ad1906908f2f5ae4e5a8ddfce73c320c2a1429ec52eafd27138b7f1cbe341c9 \
+    --hash=sha256:501a031947e3a9025ed4405a168e6ef5ae3126c59f90ce0cd6f2bfc477be31b7 \
+    --hash=sha256:5190d403f121660ce8d1d2c1bb2ef1bd05b5f68533fc5c2ea899bd15f4399b35 \
+    --hash=sha256:5498cd1645aa724a7c71c8f378eb29ebe23da2fc0d7a08071d89469bf1d2defb \
+    --hash=sha256:5cf4e27da7e3fbed4d6c3d8e797387aaad68102272f8f9752883bc32d61cb87b \
+    --hash=sha256:5e0b74767e5f8c593e8c9b5912019159ed0533c70051e9cce3e8b6aa699fcd69 \
+    --hash=sha256:5ed875a24292240029e4483f9d4a4b8a1ae08843b9c54f43fcc11e404532a8a5 \
+    --hash=sha256:5fcd34e47f6e0b794d17de1b4ff496c00986e1c83f7ab2fb8fcfe9616ff7477b \
+    --hash=sha256:5fdec68f91a0c6739b380c83b951e2c72ac0197ace422360e6d5a959d8d97b2c \
+    --hash=sha256:6344df0d5755a2c9a276d4473ae6b90647e216ab4757f8426893b5dd2ac3f369 \
+    --hash=sha256:64386e5e707d03a7e172c0701abfb7e10f0fb753ee1d773128192742712a98fd \
+    --hash=sha256:652cb6edd41e718550aad172851962662ff2681490a8a711af6a4d288dd96824 \
+    --hash=sha256:66291b10affd76d76f54fad28e22e51719ef9ba22b29e1d7d03d6777a9174198 \
+    --hash=sha256:66e1674c3ef6f541c35191caae2d429b967b99e02040f5ba928632d9a7f0f065 \
+    --hash=sha256:6adc77889b628398debc7b65c073bcb99c4a0237b248cacaf3fe8a557563ef6c \
+    --hash=sha256:79005a0d97d5ddabfeeea4cf676af11e647e41d81c9a7722a193022accdb6b7c \
+    --hash=sha256:7c6610def4f163542a622a73fb39f534f8c101d690126992300bf3207eab9764 \
+    --hash=sha256:7f047e29dcae44602496db43be01ad42fc6f1cc0d8cd6c83d342306c32270196 \
+    --hash=sha256:8098f252adfa6c80ab48096053f512f2321f0b998f98150cea9bd23d83e1467b \
+    --hash=sha256:850774a7879607d3a6f50d36d04f00ee69e7fc816450e5f7e58d7f17f1ae5c00 \
+    --hash=sha256:8d1fab6bb153a416f9aeb4b8763bc0f22a5586065f86f7664fc23339fc1c1fac \
+    --hash=sha256:8da9669d359f02c0b91ccc01cac4a67f16afec0dac22c2ad09f46bee0697eba8 \
+    --hash=sha256:8dc52c23056b9ddd46818a57b78404882310fb473d63f17b07d5c40421e47f8e \
+    --hash=sha256:9149cad251584d5fb4981be1ecde53a1ca46c891a79788c0df828d2f166bda28 \
+    --hash=sha256:93dda82c9c22deb0a405ea4dc5f2d0cda384168e466364dec6255b293923b2f3 \
+    --hash=sha256:96b533f0e99f6579b3d4d4995707cf36df9100d67e0c8303a0c55b27b5f99bc5 \
+    --hash=sha256:9c57bb8c96f6d1808c030b1687b9b5fb476abaa47f0db9c0101f5e9f394e97f4 \
+    --hash=sha256:9c7708761fccb9397fe64bbc0395abcae8c4bf7b0eac081e12b809bf47700d0b \
+    --hash=sha256:9f3bfb4965eb874431221a3ff3fdcddc7e74e3b07799e0e84ca4a0f867d449bf \
+    --hash=sha256:a33284e20b78bd4a18c8c2282d549d10bc8408a2a7ff57653c0cf0b9be0afce5 \
+    --hash=sha256:a80cb027f6b349846a3bf6d73b5e95e782175e52f22108cfa17876aaeff93702 \
+    --hash=sha256:b30236e45cf30d2b8e7b3e85881719e98507abed1011bf463a8fa23e9c3e98a8 \
+    --hash=sha256:b3bc83488de33889877a0f2543ade9f70c67d66d9ebb4ac959502e12de895788 \
+    --hash=sha256:b865addae83924361678b652338317d1bd7e79b1f4596f96b96c77a5a34b34da \
+    --hash=sha256:b8bb0864c5a28024fac8a632c443c87c5aa6f215c0b126c449ae1a150412f31d \
+    --hash=sha256:ba1cc08a7ccde2d2ec775841541641e4548226580ab850948cbfda66a1befcdc \
+    --hash=sha256:bdb2c67c6c1390b63c6ff89f210c8fd09d9a1217a465701eac7316313c915e4c \
+    --hash=sha256:c1ff362665ae507275af2853520967820d9124984e0f7466736aea23d8611fba \
+    --hash=sha256:c2514fceb77bc5e7a2f7adfaa1feb2fb311607c9cb518dbc378688ec73d8292f \
+    --hash=sha256:c3355370a2c156cffb25e876646f149d5d68f5e0a3ce86a5084dd0b64a994917 \
+    --hash=sha256:c458b6d084f9b935061bc36216e8a69a7e293a2f1e68bf956dcd9e6cbcd143f5 \
+    --hash=sha256:d0eae10f8159e8fdad514efdc92d74fd8d682c933a6dd088030f3834bc8e6b26 \
+    --hash=sha256:d76623373421df22fb4cf8817020cbb7ef15c725b9d5e45f17e189bfc384190f \
+    --hash=sha256:ebc55a14a21cb14062aa4162f906cd962b28e2e9ea38f9b4391244cd8de4ae0b \
+    --hash=sha256:eda16858a3cab07b80edaf74336ece1f986ba330fdb8ee0d6c0d68fe82bc96be \
+    --hash=sha256:ee2922902c45ae8ccada2c5b501ab86c36525b883eff4255313a253a3160861c \
+    --hash=sha256:efd7b85f94a6f21e4932043973a7ba2613b059c4a000551892ac9f1d11f5baf3 \
+    --hash=sha256:f7057c9a337546edc7973c0d3ba84ddcdf0daa14533c2065749c9075001090e6 \
+    --hash=sha256:fa160448684b4e94d80416c0fa4aac48967a969efe22931448d853ada8baf926 \
+    --hash=sha256:fc09d0aa354569bc501d4e787133afc08552722d3ab34836a80547331bb5d4a0
+    # via pre-commit
+virtualenv==21.2.4 \
+    --hash=sha256:29d21e941795206138d0f22f4e45ff7050e5da6c6472299fb7103318763861ac \
+    --hash=sha256:b294ef68192638004d72524ce7ef303e9d0cf5a44c95ce2e54a7500a6381cada
+    # via pre-commit
@@ -116,9 +116,6 @@ jobs:
    env:
      USE_DEPOT: ${{ needs.pick.outputs.is_fork != 'true' }}
      DEPOT_TOKEN: ${{ secrets.DEPOT_TOKEN }}
-      # Single source of truth for whether this preview embeds the admin portal:
-      # drives the image build-arg and the deployment comment.
-      BUILD_PORTAL: "true"

    steps:
      - name: Harden Runner
@@ -242,16 +239,14 @@ jobs:

      - name: Build and push V2 image (Depot)
        if: env.USE_DEPOT == 'true' && steps.check-image.outputs.exists == 'false'
-        uses: depot/build-push-action@98e78adca7817480b8185f474a400b451d74e287 # v1.16.0
+        uses: depot/build-push-action@5f3b3c2e5a00f0093de47f657aeaefcedff27d18 # v1.16.0
        with:
          project: ${{ vars.DEPOT_PROJECT_ID }}
          context: .
          file: ./docker/embedded/Dockerfile
          push: true
          tags: ${{ secrets.DOCKER_HUB_USERNAME }}/test:v2-${{ steps.commit-hash.outputs.app_short }}
-          build-args: |
-            VERSION_TAG=v2-alpha
-            BUILD_PORTAL=${{ env.BUILD_PORTAL }}
+          build-args: VERSION_TAG=v2-alpha
          platforms: linux/amd64

      - name: Build and push V2 image (Docker fork fallback)
@@ -264,9 +259,7 @@ jobs:
          cache-from: type=gha,scope=stirling-pdf-latest
          cache-to: type=gha,mode=max,scope=stirling-pdf-latest
          tags: ${{ secrets.DOCKER_HUB_USERNAME }}/test:v2-${{ steps.commit-hash.outputs.app_short }}
-          build-args: |
-            VERSION_TAG=v2-alpha
-            BUILD_PORTAL=${{ env.BUILD_PORTAL }}
+          build-args: VERSION_TAG=v2-alpha
          platforms: linux/amd64

      - name: Set up SSH
@@ -297,12 +290,10 @@ jobs:
                - /stirling/V2-PR-${{ needs.check-pr.outputs.pr_number }}/storage:/storage:rw
              environment:
                DISABLE_ADDITIONAL_FEATURES: "false"
-                POLICIES_ENABLED: "true"
-                STIRLING_BILLING_ACCOUNT_LINK_ENABLED: "true"
                SECURITY_ENABLELOGIN: "true"
                SECURITY_INITIALLOGIN_USERNAME: "${{ secrets.TEST_LOGIN_USERNAME }}"
                SECURITY_INITIALLOGIN_PASSWORD: "${{ secrets.TEST_LOGIN_PASSWORD }}"
-                SYSTEM_DEFAULTLOCALE: en-US
+                SYSTEM_DEFAULTLOCALE: en-GB
                UI_APPNAME: "Stirling-PDF V2 PR#${{ needs.check-pr.outputs.pr_number }}"
                UI_HOMEDESCRIPTION: "V2 PR#${{ needs.check-pr.outputs.pr_number }} - Embedded Architecture"
                UI_APPNAMENAVBAR: "V2 PR#${{ needs.check-pr.outputs.pr_number }}"
@@ -342,70 +333,9 @@ jobs:
          # Set port for output
          echo "v2_port=${V2_PORT}" >> $GITHUB_OUTPUT

-      # ---- Storybook preview (only when this PR touches stories/.storybook) ----
-      # Runs inside the same approved-contributor-gated deploy job, so it deploys
-      # under the exact same access rules as the app preview.
-      - name: Detect Storybook changes
-        id: sb-changes
-        uses: dorny/paths-filter@fbd0ab8f3e69293af611ebaee6363fc25e6d187d # v4.0.1
-        with:
-          list-files: json
-          filters: |
-            storybook:
-              - 'frontend/**/*.stories.@(ts|tsx|mdx)'
-              - 'frontend/**/*.mdx'
-              - 'frontend/.storybook/**'
-
-      - name: Set up Node.js for Storybook
-        if: steps.sb-changes.outputs.storybook == 'true'
-        uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
-        with:
-          node-version: "22"
-          cache: "npm"
-          cache-dependency-path: frontend/package-lock.json
-
-      - name: Install Task for Storybook
-        if: steps.sb-changes.outputs.storybook == 'true'
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
-
-      - name: Build and deploy Storybook
-        id: storybook
-        if: steps.sb-changes.outputs.storybook == 'true'
-        env:
-          VPS_HOST: ${{ secrets.NEW_VPS_HOST }}
-          VPS_USER: ${{ secrets.NEW_VPS_USERNAME }}
-        run: |
-          set -euo pipefail
-          # `prepare` generates the icon set stories import (not committed).
-          task frontend:prepare
-          task frontend:storybook:build
-          PR=${{ needs.check-pr.outputs.pr_number }}
-          # Served at the ROOT of its own port so Storybook's global MSW worker
-          # (/mockServiceWorker.js) resolves. Port = PR + 20000 (bijective, offset
-          # from the app preview's bare-PR-number port).
-          SB_PORT=$((PR + 20000))
-          DIR=/stirling/SB-PR-$PR
-          tar czf storybook.tgz -C frontend/storybook-static .
-          scp -i ../private.key -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null \
-            storybook.tgz "$VPS_USER@$VPS_HOST:/tmp/storybook-$PR.tgz"
-          ssh -i ../private.key -o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -T \
-            "$VPS_USER@$VPS_HOST" << ENDSSH
-            set -e
-            rm -rf "$DIR" && mkdir -p "$DIR"
-            tar xzf /tmp/storybook-$PR.tgz -C "$DIR"
-            rm -f /tmp/storybook-$PR.tgz
-            docker rm -f storybook-pr-$PR 2>/dev/null || true
-            docker run -d --name storybook-pr-$PR --restart unless-stopped \
-              -p $SB_PORT:80 -v "$DIR":/usr/share/nginx/html:ro nginx:alpine
-          ENDSSH
-          echo "url=http://$VPS_HOST:$SB_PORT/" >> "$GITHUB_OUTPUT"
-
      - name: Post V2 deployment URL to PR
        if: success()
        uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
-        env:
-          SB_URL: ${{ steps.storybook.outputs.url }}
-          SB_FILES: ${{ steps.sb-changes.outputs.storybook_files }}
        with:
          github-token: ${{ steps.setup-bot.outputs.token }}
          script: |
@@ -429,40 +359,12 @@ jobs:
            }

            const deploymentUrl = `http://${{ secrets.NEW_VPS_HOST }}:${v2Port}`;
-
-            // Only mention the portal when this image actually embeds it.
-            // Use the direct IP URL - the SSL hostname isn't supported yet.
-            const withPortal = "${{ env.BUILD_PORTAL }}" === "true";
-            const portalNote = withPortal
-              ? `🧩 **Admin portal** included - try it at [${deploymentUrl}/portal](${deploymentUrl}/portal).\n\n`
-              : ``;
-
-            // Storybook preview: only present when this PR changed stories/config.
-            const sbUrl = process.env.SB_URL;
-            let storybookNote = "";
-            if (sbUrl) {
-              const files = JSON.parse(process.env.SB_FILES || "[]");
-              const stories = files.filter((f) => /\.stories\.(ts|tsx|mdx)$/.test(f));
-              const config = files.filter((f) => f.startsWith("frontend/.storybook/"));
-              const shorten = (f) =>
-                f.replace(/^frontend\/editor\/src\//, "").replace(/^frontend\//, "");
-              const storyList = stories.map((f) => `- \`${shorten(f)}\``).join("\n");
-              const configList = config.map((f) => `- \`${shorten(f)}\``).join("\n");
-              const summary =
-                `${stories.length} stor${stories.length === 1 ? "y" : "ies"} changed` +
-                (config.length ? ` (+${config.length} config file${config.length === 1 ? "" : "s"})` : "");
-              storybookNote =
-                `📚 **Storybook:** [${sbUrl}](${sbUrl})\n\n` +
-                `<details>\n<summary>${summary}</summary>\n\n` +
-                (storyList ? `**Stories**\n${storyList}\n\n` : "") +
-                (configList ? `**Config**\n${configList}\n` : "") +
-                `</details>\n\n`;
-            }
+            const httpsUrl = `https://${v2Port}.ssl.stirlingpdf.cloud`;

            const commentBody = `## 🚀 V2 Auto-Deployment Complete!\n\n` +
+                              `Your V2 PR with embedded architecture has been deployed!\n\n` +
                              `🔗 **Direct Test URL (non-SSL)** [${deploymentUrl}](${deploymentUrl})\n\n` +
-                              portalNote +
-                              storybookNote +
+                              `🔐 **Secure HTTPS URL**: [${httpsUrl}](${httpsUrl})\n\n` +
                              `_This deployment will be automatically cleaned up when the PR is closed._\n\n` +
                              `🔄 **Auto-deployed** for approved V2 contributors.`;

@@ -558,11 +460,7 @@ jobs:
            else
              echo "V2 PR directory not found, nothing to clean up"
            fi
-
-            # Remove this PR's Storybook preview (container + files), if any.
-            docker rm -f storybook-pr-${{ github.event.pull_request.number }} 2>/dev/null || true
-            rm -rf /stirling/SB-PR-${{ github.event.pull_request.number }}
-
+            
            # Clean up old unused images (older than 2 weeks) but keep recent ones for reuse
            docker image prune -af --filter "until=336h" --filter "label!=keep=true" || true

@@ -222,10 +222,10 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Run Gradle Command
        run: |
          if [ "${{ needs.check-comment.outputs.disable_security }}" == "true" ]; then
@@ -256,7 +256,7 @@ jobs:

      - name: Build and push PR-specific image (Depot)
        if: env.USE_DEPOT == 'true'
-        uses: depot/build-push-action@98e78adca7817480b8185f474a400b451d74e287 # v1.16.0
+        uses: depot/build-push-action@5f3b3c2e5a00f0093de47f657aeaefcedff27d18 # v1.16.0
        with:
          project: ${{ vars.DEPOT_PROJECT_ID }}
          context: .
@@ -285,7 +285,7 @@ jobs:

      - name: Build and push engine image (Depot)
        if: env.USE_DEPOT == 'true' && needs.check-comment.outputs.enable_prototypes == 'true'
-        uses: depot/build-push-action@98e78adca7817480b8185f474a400b451d74e287 # v1.16.0
+        uses: depot/build-push-action@5f3b3c2e5a00f0093de47f657aeaefcedff27d18 # v1.16.0
        with:
          project: ${{ vars.DEPOT_PROJECT_ID }}
          context: ./engine
@@ -388,7 +388,7 @@ jobs:
              environment:
                DISABLE_ADDITIONAL_FEATURES: "${DISABLE_ADDITIONAL_FEATURES}"
                SECURITY_ENABLELOGIN: "${LOGIN_SECURITY}"
-                SYSTEM_DEFAULTLOCALE: en-US
+                SYSTEM_DEFAULTLOCALE: en-GB
                UI_APPNAME: "Stirling-PDF PR#${PR_NUMBER}"
                UI_HOMEDESCRIPTION: "PR#${PR_NUMBER} for Stirling-PDF Latest"
                UI_APPNAMENAVBAR: "PR#${PR_NUMBER}"
@@ -1,9 +1,9 @@
 name: AI Engine CI

-# Runs the engine quality gate (lint, type-check, format-check, tests). Called
-# from build.yml on PRs and merge_group; also runs directly on push to main as
-# a post-merge safety net. Freshness of the generated tool_models.py is checked
-# by the shared check-generated-models workflow.
+# Validates the Python AI engine: regenerates tool models and runs the
+# engine quality gate (lint, type-check, format-check, tests). Called from
+# build.yml on PRs and merge_group; also runs directly on push to main as
+# a post-merge safety net.
 on:
  workflow_call:
  push:
@@ -30,12 +30,107 @@ jobs:
        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2

      - name: Install uv
-        uses: astral-sh/setup-uv@fac544c07dec837d0ccb6301d7b5580bf5edae39 # v8.2.0
+        uses: astral-sh/setup-uv@08807647e7069bb48b6ef5acd8ec9567f424441b # v8.1.0
        with:
          enable-cache: true

+      - name: Set up JDK 25
+        uses: actions/setup-java@be666c2fcd27ec809703dec50e508c2fdc7f6654 # v5.2.0
+        with:
+          java-version: "25"
+          distribution: "temurin"
+
+      - name: Setup Gradle
+        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
+        with:
+          gradle-version: 9.3.1
+
      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
+
+      - name: Regenerate tool models
+        run: task engine:tool-models
+
+      - name: Verify tool models are up to date
+        id: tool-models-check
+        continue-on-error: true
+        run: git diff --exit-code engine/src/stirling/models/tool_models.py
+
+      - name: Comment on tool models check failure
+        # Only post a comment on PRs. github-script's PR helpers need an
+        # issue/PR number, which doesn't exist on merge_group runs.
+        if: steps.tool-models-check.outcome == 'failure' && github.event_name == 'pull_request'
+        continue-on-error: true
+        uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
+        with:
+          script: |
+            const marker = '<!-- tool-models-check -->';
+            const body = [
+              marker,
+              '### Tool Models Check Failed',
+              '',
+              'The generated `engine/src/stirling/models/tool_models.py` is out of date with the Java OpenAPI spec and will need to be regenerated before it can be merged in.',
+              '',
+              'Run `task engine:tool-models` to regenerate, then commit the updated file.',
+            ].join('\n');
+            const { data: comments } = await github.rest.issues.listComments({
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              issue_number: context.issue.number,
+            });
+            const existing = comments.find(c => c.body.includes(marker));
+            if (existing) {
+              await github.rest.issues.updateComment({
+                owner: context.repo.owner,
+                repo: context.repo.repo,
+                comment_id: existing.id,
+                body,
+              });
+            } else {
+              await github.rest.issues.createComment({
+                owner: context.repo.owner,
+                repo: context.repo.repo,
+                issue_number: context.issue.number,
+                body,
+              });
+            }
+
+      - name: Fail if tool models check failed
+        if: steps.tool-models-check.outcome == 'failure'
+        run: |
+          echo "============================================"
+          echo "  Tool Models Check Failed"
+          echo "============================================"
+          echo ""
+          echo "The generated engine/src/stirling/models/tool_models.py"
+          echo "is out of date with the Java OpenAPI spec and will"
+          echo "need to be regenerated before it can be merged in."
+          echo ""
+          echo "Run 'task engine:tool-models' to regenerate, then"
+          echo "commit the updated file."
+          echo "============================================"
+          exit 1
+
+      - name: Remove tool models check comment on success
+        if: steps.tool-models-check.outcome == 'success' && github.event_name == 'pull_request'
+        continue-on-error: true
+        uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
+        with:
+          script: |
+            const marker = '<!-- tool-models-check -->';
+            const { data: comments } = await github.rest.issues.listComments({
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              issue_number: context.issue.number,
+            });
+            const existing = comments.find(c => c.body.includes(marker));
+            if (existing) {
+              await github.rest.issues.deleteComment({
+                owner: context.repo.owner,
+                repo: context.repo.repo,
+                comment_id: existing.id,
+              });
+            }

      - name: Quality-check engine
        id: engine-check
@@ -58,11 +58,11 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1
          cache-disabled: true

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Check Java formatting (Spotless)
        # Runs once per matrix combination - pick the cheapest leg
        # (core - no proprietary, no saas) so we don't wait for the
@@ -2,7 +2,7 @@ name: Enterprise E2E (Playwright)

 # Enterprise Playwright suite — exercises premium-key gated features (audit,
 # teams, analytics) plus full OAuth + SAML logins via the Keycloak compose
-# stacks under testing/compose. Slow and secret-gated, so it runs in four
+# stacks under testing/compose. Slow and secret-gated, so it runs in three
 # situations:
 #
 #   - PRs that touch proprietary / premium / SSO compose / enterprise tests
@@ -12,6 +12,8 @@ name: Enterprise E2E (Playwright)
 #   - on a nightly cron schedule (catches Keycloak image drift, license
 #     expiry, upstream proprietary changes),
 #   - manual workflow_dispatch.
+#
+# Auto-skipped when secrets.PREMIUM_KEY_ENTERPRISE is missing (forks, dependabot).

 on:
  workflow_call:
@@ -50,10 +52,6 @@ jobs:

  playwright-e2e-enterprise:
    needs: pick
-    # Skip on fork PRs / untrusted authors: they have no PREMIUM_KEY_ENTERPRISE
-    # (nor DEPOT_TOKEN), so the suite can't boot premium and would fail. See the
-    # header comment. GitHub reports the skipped reusable workflow as success.
-    if: needs.pick.outputs.is_fork != 'true'
    runs-on: ${{ needs.pick.outputs.is_fork == 'true' && 'ubuntu-latest' || format('depot-ubuntu-24.04-{0}', inputs.depot_cores || '8') }}
    timeout-minutes: 45
    env:
@@ -80,7 +78,7 @@ jobs:
          cache: "npm"
          cache-dependency-path: frontend/package-lock.json
      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Install Playwright (chromium only)
        run: task e2e:install -- chromium
      - name: Build frontend (needed for playwright's vite preview webServer)
@@ -167,8 +165,6 @@ jobs:
          wait_for_backend
      - name: Run enterprise OAuth Playwright tests
        id: oauth-tests
-        env:
-          PLAYWRIGHT_JSON_OUTPUT_FILE: ${{ github.workspace }}/frontend/playwright-report/results-oauth.json
        run: task e2e:enterprise -- --grep "OAuth"
      - name: Stop backend + tear down OAuth Keycloak
        if: always()
@@ -242,8 +238,6 @@ jobs:
          wait_for_backend
      - name: Run enterprise SAML Playwright tests
        id: saml-tests
-        env:
-          PLAYWRIGHT_JSON_OUTPUT_FILE: ${{ github.workspace }}/frontend/playwright-report/results-saml.json
        run: task e2e:enterprise -- --grep "SAML"
      - name: Stop backend + tear down SAML Keycloak
        if: always()
@@ -274,8 +268,6 @@ jobs:
          wait_for_backend
      - name: Run enterprise feature Playwright tests
        id: feature-tests
-        env:
-          PLAYWRIGHT_JSON_OUTPUT_FILE: ${{ github.workspace }}/frontend/playwright-report/results-feature.json
        run: task e2e:enterprise -- --grep "Enterprise license"
      - name: Print backend log on failure
        if: failure()
@@ -288,23 +280,10 @@ jobs:
        run: |
          source /tmp/helpers.sh
          stop_backend
-      - name: Flag flaky tests
-        # Runs regardless of the test outcomes: a flaky test (passed on retry)
-        # leaves its step green, so this is the only place it surfaces. Merges
-        # all three phase reports (some may be absent if an earlier phase hard-
-        # failed and skipped the rest). Emits ::warning:: annotations + a job
-        # summary; never fails the job.
-        if: always()
-        working-directory: frontend
-        run: >
-          npx tsx editor/scripts/report-flaky-tests.mts
-          "${{ github.workspace }}/frontend/playwright-report/results-oauth.json"
-          "${{ github.workspace }}/frontend/playwright-report/results-saml.json"
-          "${{ github.workspace }}/frontend/playwright-report/results-feature.json"
      - name: Upload Playwright report
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: playwright-report-enterprise-${{ github.run_id }}
-          path: frontend/playwright-report/
+          path: frontend/editor/playwright-report/
          retention-days: 7
@@ -43,7 +43,6 @@ jobs:
      docker-base: ${{ steps.changes.outputs.docker-base }}
      tauri: ${{ steps.changes.outputs.tauri }}
      engine: ${{ steps.changes.outputs.engine }}
-      generated-models: ${{ steps.changes.outputs.generated-models }}
      proprietary: ${{ steps.changes.outputs.proprietary }}
    steps:
      - name: Harden the runner (Audit all outbound calls)
@@ -172,20 +171,6 @@ jobs:
    uses: ./.github/workflows/ai-engine.yml
    secrets: inherit

-  # The generated frontend types and engine tool models are both derived from
-  # the Java OpenAPI spec. This job regenerates and diffs them; it boots the
-  # backend, so it is gated on the narrow generated-models filter (spec source,
-  # generators, generated files, generation tasks) rather than the broad
-  # frontend filter, so a CSS-only PR does not pay for a backend build.
-  generated-models:
-    if: needs.files-changed.outputs.generated-models == 'true'
-    needs: [files-changed]
-    permissions:
-      contents: read
-      pull-requests: write
-    uses: ./.github/workflows/check-generated-models.yml
-    secrets: inherit
-
  pre-commit:
    needs: [files-changed]
    permissions:
@@ -217,9 +202,6 @@ jobs:
      contents: read
    uses: ./.github/workflows/coverage-aggregate.yml
    secrets: inherit
-    with:
-      frontend-validation-result: ${{ needs.frontend-validation.result }}
-      playwright-e2e-live-result: ${{ needs.playwright-e2e-live.result }}

  # Single status check that branch protection should mark as required.
  # Succeeds when every upstream job is either `success` or `skipped` (path-
@@ -243,7 +225,6 @@ jobs:
      - test-build-docker-images
      - tauri-build
      - ai-engine
-      - generated-models
      - pre-commit
      - dependency-review
    runs-on: ubuntu-latest
@@ -269,7 +250,6 @@ jobs:
            test-build-docker-images=${{ needs.test-build-docker-images.result }}
            tauri-build=${{ needs.tauri-build.result }}
            ai-engine=${{ needs.ai-engine.result }}
-            generated-models=${{ needs.generated-models.result }}
            pre-commit=${{ needs.pre-commit.result }}
            dependency-review=${{ needs.dependency-review.result }}
        run: |
@@ -1,148 +0,0 @@
-name: Check generated models
-
-# Verifies the committed generated API models are still in sync with the Java
-# OpenAPI spec: the frontend tool API types
-# (frontend/editor/src/core/types/toolApiTypes.ts) and the engine tool
-# models (engine/src/stirling/models/tool_models.py). Regenerates both with the
-# single top-level `task tool-models` and fails if either committed file is
-# out of date. Called from build.yml when the backend Java, frontend, or engine
-# changes; also runs on push to main as a post-merge safety net.
-on:
-  workflow_call:
-  push:
-    branches: [main]
-
-permissions:
-  contents: read
-
-jobs:
-  generated-models:
-    runs-on: ubuntu-latest
-    permissions:
-      contents: read
-      pull-requests: write
-    env:
-      DEPOT_TOKEN: ${{ secrets.DEPOT_TOKEN }}
-    steps:
-      - name: Harden the runner (Audit all outbound calls)
-        uses: step-security/harden-runner@ab7a9404c0f3da075243ca237b5fac12c98deaa5 # v2.19.3
-        with:
-          egress-policy: audit
-
-      - name: Checkout code
-        uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
-
-      - name: Install uv
-        uses: astral-sh/setup-uv@fac544c07dec837d0ccb6301d7b5580bf5edae39 # v8.2.0
-        with:
-          enable-cache: true
-
-      - name: Set up JDK 25
-        uses: actions/setup-java@be666c2fcd27ec809703dec50e508c2fdc7f6654 # v5.2.0
-        with:
-          java-version: "25"
-          distribution: "temurin"
-
-      - name: Setup Gradle
-        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
-        with:
-          gradle-version: 9.6.0
-
-      - name: Set up Node
-        uses: actions/setup-node@48b55a011bda9f5d6aeb4c2d9c7362e8dae4041e # v6.4.0
-        with:
-          node-version: "22"
-          cache: "npm"
-          cache-dependency-path: frontend/package-lock.json
-
-      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
-
-      # Rebuilds the OpenAPI spec from the current Java and regenerates both the
-      # frontend types and the engine tool models from it.
-      - name: Regenerate generated models
-        run: task tool-models
-
-      - name: Verify generated models are up to date
-        id: models-check
-        continue-on-error: true
-        run: |
-          git diff --exit-code \
-            frontend/editor/src/core/types/toolApiTypes.ts \
-            engine/src/stirling/models/tool_models.py
-
-      - name: Comment on generated models check failure
-        # Only post a comment on PRs. github-script's PR helpers need an
-        # issue/PR number, which doesn't exist on merge_group runs.
-        if: steps.models-check.outcome == 'failure' && github.event_name == 'pull_request'
-        continue-on-error: true
-        uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
-        with:
-          script: |
-            const marker = '<!-- generated-models-check -->';
-            const body = [
-              marker,
-              '### Generated Models Check Failed',
-              '',
-              'The generated `frontend/editor/src/core/types/toolApiTypes.ts` and/or `engine/src/stirling/models/tool_models.py` are out of date with the Java OpenAPI spec and will need to be regenerated before they can be merged in.',
-              '',
-              'Run `task tool-models` to regenerate both, then commit the updated files.',
-            ].join('\n');
-            const { data: comments } = await github.rest.issues.listComments({
-              owner: context.repo.owner,
-              repo: context.repo.repo,
-              issue_number: context.issue.number,
-            });
-            const existing = comments.find(c => c.body.includes(marker));
-            if (existing) {
-              await github.rest.issues.updateComment({
-                owner: context.repo.owner,
-                repo: context.repo.repo,
-                comment_id: existing.id,
-                body,
-              });
-            } else {
-              await github.rest.issues.createComment({
-                owner: context.repo.owner,
-                repo: context.repo.repo,
-                issue_number: context.issue.number,
-                body,
-              });
-            }
-
-      - name: Fail if generated models check failed
-        if: steps.models-check.outcome == 'failure'
-        run: |
-          echo "============================================"
-          echo "  Generated Models Check Failed"
-          echo "============================================"
-          echo ""
-          echo "The generated frontend API types and/or engine tool"
-          echo "models are out of date with the Java OpenAPI spec and"
-          echo "will need to be regenerated before they can be merged in."
-          echo ""
-          echo "Run 'task tool-models' to regenerate both, then"
-          echo "commit the updated files."
-          echo "============================================"
-          exit 1
-
-      - name: Remove generated models check comment on success
-        if: steps.models-check.outcome == 'success' && github.event_name == 'pull_request'
-        continue-on-error: true
-        uses: actions/github-script@3a2844b7e9c422d3c10d287c895573f7108da1b3 # v9.0.0
-        with:
-          script: |
-            const marker = '<!-- generated-models-check -->';
-            const { data: comments } = await github.rest.issues.listComments({
-              owner: context.repo.owner,
-              repo: context.repo.repo,
-              issue_number: context.issue.number,
-            });
-            const existing = comments.find(c => c.body.includes(marker));
-            if (existing) {
-              await github.rest.issues.deleteComment({
-                owner: context.repo.owner,
-                repo: context.repo.repo,
-                comment_id: existing.id,
-              });
-            }
@@ -40,11 +40,11 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1
          cache-disabled: true

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Check licenses for compatibility
        run: task backend:licenses:check
        env:
@@ -45,11 +45,11 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1
          cache-disabled: true

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Generate OpenAPI documentation
        run: task backend:swagger
        env:
@@ -166,16 +166,16 @@ jobs:

            // Determine reference file
            let referenceFilePath;
-            if (changedFiles.includes("frontend/editor/public/locales/en-US/translation.toml")) {
+            if (changedFiles.includes("frontend/editor/public/locales/en-GB/translation.toml")) {
              console.log("Using PR branch reference file.");
              const { data: fileContent } = await github.rest.repos.getContent({
                owner: prRepoOwner,
                repo: prRepoName,
-                path: "frontend/editor/public/locales/en-US/translation.toml",
+                path: "frontend/editor/public/locales/en-GB/translation.toml",
                ref: branch,
              });

-              referenceFilePath = "pr-branch-translation-en-US.toml";
+              referenceFilePath = "pr-branch-translation-en-GB.toml";
              const content = Buffer.from(fileContent.content, "base64").toString("utf-8");
              fs.writeFileSync(referenceFilePath, content);
            } else {
@@ -183,11 +183,11 @@ jobs:
              const { data: fileContent } = await github.rest.repos.getContent({
                owner: repoOwner,
                repo: repoName,
-                path: "frontend/editor/public/locales/en-US/translation.toml",
+                path: "frontend/editor/public/locales/en-GB/translation.toml",
                ref: "main",
              });

-              referenceFilePath = "main-branch-translation-en-US.toml";
+              referenceFilePath = "main-branch-translation-en-GB.toml";
              const content = Buffer.from(fileContent.content, "base64").toString("utf-8");
              fs.writeFileSync(referenceFilePath, content);
            }
@@ -293,6 +293,6 @@ jobs:
        run: |
          echo "Cleaning up temporary files..."
          rm -rf pr-branch
-          rm -f pr-branch-translation-en-US.toml main-branch-translation-en-US.toml changed_files.txt result.txt
+          rm -f pr-branch-translation-en-GB.toml main-branch-translation-en-GB.toml changed_files.txt result.txt
          echo "Cleanup complete."
        continue-on-error: true # Ensure cleanup runs even if previous steps fail
@@ -13,17 +13,6 @@ name: Aggregate backend coverage
 #     producers themselves
 on:
  workflow_call:
-    inputs:
-      frontend-validation-result:
-        description: Result of the frontend-validation producer job
-        required: false
-        type: string
-        default: skipped
-      playwright-e2e-live-result:
-        description: Result of the playwright-e2e-live producer job
-        required: false
-        type: string
-        default: skipped

 permissions:
  contents: read
@@ -62,7 +51,7 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1
          cache-disabled: true

      - name: Set up Python
@@ -207,9 +196,9 @@ jobs:
      # --------------------------------------------------------------
      - name: Download vitest coverage artifact
        # frontend-validation uploads as `frontend-coverage`. Tolerate
-        # absence on backend-only runs by skipping the download entirely
-        # when the producer job was not part of this workflow run.
-        if: inputs.frontend-validation-result == 'success'
+        # absence so a backend-only PR still produces the matrix with
+        # just backend rows populated.
+        if: always()
        uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v6.0.0
        with:
          name: frontend-coverage
@@ -217,12 +206,12 @@ jobs:
        continue-on-error: true

      - name: Download Playwright frontend coverage artifact
-        # e2e-live uploads the artifact with a stable name. Skip the
-        # download entirely when the producer job did not run.
-        if: inputs.playwright-e2e-live-result == 'success'
+        # e2e-live uploads as `playwright-frontend-coverage-<run_id>`.
+        # Same tolerance as vitest - matrix script handles missing inputs.
+        if: always()
        uses: actions/download-artifact@d3f86a106a0bac45b974a628896c90dbdf5c8093 # v6.0.0
        with:
-          name: playwright-frontend-coverage
+          name: playwright-frontend-coverage-${{ github.run_id }}
          path: matrix-inputs/playwright/
        continue-on-error: true

@@ -48,7 +48,7 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1
          cache-disabled: true

      # No `-PnoSpotless` here yet because the upstream cache layer matches the
@@ -107,7 +107,7 @@ jobs:

      - name: Build and push frontend image (Depot)
        if: env.USE_DEPOT == 'true' && steps.check-frontend.outputs.exists == 'false'
-        uses: depot/build-push-action@98e78adca7817480b8185f474a400b451d74e287 # v1.16.0
+        uses: depot/build-push-action@5f3b3c2e5a00f0093de47f657aeaefcedff27d18 # v1.16.0
        with:
          project: ${{ vars.DEPOT_PROJECT_ID }}
          context: .
@@ -136,7 +136,7 @@ jobs:

      - name: Build and push backend image (Depot)
        if: env.USE_DEPOT == 'true' && steps.check-backend.outputs.exists == 'false'
-        uses: depot/build-push-action@98e78adca7817480b8185f474a400b451d74e287 # v1.16.0
+        uses: depot/build-push-action@5f3b3c2e5a00f0093de47f657aeaefcedff27d18 # v1.16.0
        with:
          project: ${{ vars.DEPOT_PROJECT_ID }}
          context: .
@@ -188,7 +188,7 @@ jobs:
              environment:
                DISABLE_ADDITIONAL_FEATURES: "true"
                SECURITY_ENABLELOGIN: "false"
-                SYSTEM_DEFAULTLOCALE: en-US
+                SYSTEM_DEFAULTLOCALE: en-GB
                UI_APPNAME: "Stirling-PDF V2"
                UI_HOMEDESCRIPTION: "V2 Frontend/Backend Split"
                UI_APPNAMENAVBAR: "V2 Deployment"
@@ -61,22 +61,14 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1
          cache-disabled: true

-      # When the PR changes the base image, test.sh builds it locally
-      # (stirling-pdf-base:local) into the daemon image store. A buildx
-      # container builder can't see that store, so skip it here and let
-      # `docker buildx build` fall back to the default docker driver, which
-      # resolves the local base. The gha cache backend is also skipped (its
-      # runtime token isn't exposed) since the docker driver can't use it.
      - name: Set up Docker Buildx
-        if: inputs.docker-base-changed != 'true'
        uses: docker/setup-buildx-action@4d04d5d9486b7bd6fa91e7baf45bbb4f8b9deedd # v4.0.0

      # Expose ACTIONS_RUNTIME_TOKEN / ACTIONS_RESULTS_URL for docker buildx type=gha cache backend.
      - name: Expose GitHub runtime for Buildx cache
-        if: inputs.docker-base-changed != 'true'
        uses: crazy-max/ghaction-github-runtime@04d248b84655b509d8c44dc1d6f990c879747487 # v4.0.0

      - name: Install Docker Compose
@@ -42,7 +42,7 @@ jobs:
          cache: "npm"
          cache-dependency-path: frontend/package-lock.json
      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Install Playwright (chromium only)
        run: task e2e:install -- chromium
      - name: Build frontend (production bundle for vite preview)
@@ -62,17 +62,7 @@ jobs:
          # .test-state/playwright/coverage-pw/ for the post-process step
          # to aggregate. Chromium-only - other engines silently skip.
          PW_COVERAGE: "1"
-          PLAYWRIGHT_JSON_OUTPUT_FILE: ${{ github.workspace }}/frontend/playwright-report/results.json
        run: task e2e:live
-      - name: Flag flaky tests
-        # Runs regardless of the test outcome: a flaky test (passed on retry)
-        # leaves the step green, so this is the only place it surfaces. Emits
-        # ::warning:: annotations + a job summary; never fails the job.
-        if: always()
-        working-directory: frontend
-        run: npx tsx editor/scripts/report-flaky-tests.mts "$PLAYWRIGHT_JSON_OUTPUT_FILE"
-        env:
-          PLAYWRIGHT_JSON_OUTPUT_FILE: ${{ github.workspace }}/frontend/playwright-report/results.json
      - name: Generate JaCoCo report from e2e:live .exec
        if: always()
        id: live-coverage
@@ -179,7 +169,7 @@ jobs:
        if: always() && steps.pw-frontend-coverage.outputs.summary == 'true'
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
-          name: playwright-frontend-coverage
+          name: playwright-frontend-coverage-${{ github.run_id }}
          path: |
            .test-state/playwright/coverage-pw-summary/
            .test-state/playwright/coverage-pw/
@@ -198,30 +188,10 @@ jobs:
          name: backend-log-live-${{ github.run_id }}
          path: .test-state/playwright/backend.log
          retention-days: 7
-      - name: List Playwright output locations (debug)
-        if: always()
-        run: |
-          echo "::group::Playwright output dirs"
-          # Playwright anchors its default outputDir + HTML report to the
-          # nearest package.json, which is frontend/ (frontend/editor has
-          # none), so artifacts land under frontend/, not frontend/editor/.
-          ls -la frontend/playwright-report 2>/dev/null \
-            || echo "no playwright-report at frontend/"
-          ls -la frontend/test-results 2>/dev/null \
-            || echo "no test-results at frontend/"
-          find . -name node_modules -prune -o -name 'trace.zip' -print 2>/dev/null || true
-          echo "::endgroup::"
-      - name: Upload Playwright report + traces
+      - name: Upload Playwright report
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: playwright-report-live-${{ github.run_id }}
-          # test-results/ holds the per-test trace.zip (with browser console
-          # logs) + screenshots/video; playwright-report/ is the HTML report.
-          # Both live under frontend/ (Playwright anchors them to the nearest
-          # package.json, which is frontend/; frontend/editor has none).
-          path: |
-            frontend/playwright-report/
-            frontend/test-results/
+          path: frontend/editor/playwright-report/
          retention-days: 7
-          if-no-files-found: warn
@@ -36,7 +36,7 @@ jobs:
          cache: "npm"
          cache-dependency-path: frontend/package-lock.json
      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Install Playwright (chromium only)
        run: task e2e:install -- chromium
      - name: Build frontend (production bundle for vite preview)
@@ -44,22 +44,11 @@ jobs:
          VITE_BUILD_FOR_PREVIEW: "1"
        run: task frontend:build
      - name: Run stubbed E2E tests (chromium)
-        env:
-          PLAYWRIGHT_JSON_OUTPUT_FILE: ${{ github.workspace }}/frontend/playwright-report/results.json
        run: task e2e:stubbed -- --workers=3
-      - name: Flag flaky tests
-        # Runs regardless of the test outcome: a flaky test (passed on retry)
-        # leaves the step green, so this is the only place it surfaces. Emits
-        # ::warning:: annotations + a job summary; never fails the job.
-        if: always()
-        working-directory: frontend
-        run: npx tsx editor/scripts/report-flaky-tests.mts "$PLAYWRIGHT_JSON_OUTPUT_FILE"
-        env:
-          PLAYWRIGHT_JSON_OUTPUT_FILE: ${{ github.workspace }}/frontend/playwright-report/results.json
      - name: Upload Playwright report
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
          name: playwright-report-stubbed-${{ github.run_id }}
-          path: frontend/playwright-report/
+          path: frontend/editor/playwright-report/
          retention-days: 7
@@ -97,14 +97,7 @@ jobs:
        run: npm ci --ignore-scripts --audit=false --fund=false

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
-
-      - name: Generate frontend license report (Push only)
-        if: github.event_name == 'push'
-        env:
-          PR_IS_FORK: "false"
-        run: task frontend:licenses:generate
-
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Generate frontend license report (internal PR)
        if: github.event_name == 'pull_request' && github.event.pull_request.head.repo.fork == false
        env:
@@ -356,11 +349,10 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
-
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Check licenses and generate report
        id: license-check
        run: task backend:licenses:generate || echo "LICENSE_CHECK_FAILED=true" >> $GITHUB_ENV
@@ -31,7 +31,7 @@ jobs:
          cache: "npm"
          cache-dependency-path: frontend/package-lock.json
      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Quality-check frontend
        id: frontend-check
        run: task frontend:check:all
@@ -73,10 +73,10 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Get version number
        id: versionNumber
        run: |
@@ -148,7 +148,7 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1

      - name: Setup Node.js
        if: matrix.variant.build_frontend == true
@@ -159,7 +159,7 @@ jobs:
          cache-dependency-path: frontend/package-lock.json

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0

      - name: Build JAR
        run: ./gradlew build ${{ matrix.variant.build_frontend && '-PbuildWithFrontend=true' || '' }} -x spotlessApply -x spotlessCheck -x test -x sonarqube
@@ -252,10 +252,10 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0

      # Build the universal JRE before desktop:prepare so the jlink:runtime
      # task short-circuits on its `test -d runtime/jre` status check.
@@ -37,15 +37,10 @@ jobs:
          cache-dependency-path: frontend/package-lock.json

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Install all Playwright browsers
        run: task e2e:install

-      - name: Build frontend (production bundle for vite preview)
-        env:
-          VITE_BUILD_FOR_PREVIEW: "1"
-        run: task frontend:build
-
      - name: Run E2E tests (all browsers)
        run: task e2e:cross-browser

@@ -53,19 +48,6 @@ jobs:
        if: always()
        uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
        with:
-          name: playwright-report-nightly-${{ github.run_id }}
-          path: frontend/playwright-report/
+          name: playwright-nightly-${{ github.run_id }}
+          path: frontend/editor/playwright-report/
          retention-days: 14
-
-  # Builds all desktop platforms on a schedule so the Rust dependency cache is
-  # written on main, where PR and merge-queue tauri builds can restore it.
-  warm-tauri-cache:
-    name: Warm Tauri Rust cache
-    permissions:
-      contents: read
-      pull-requests: write
-    uses: ./.github/workflows/tauri-build.yml
-    with:
-      platform: all
-      sign: false
-    secrets: inherit
@@ -1,7 +1,8 @@
 name: Pre-commit

-# Runs the repo-wide lint/format/secret checks via `task pre-commit`.
-# Called from build.yml on PRs and merge_group; also runnable on demand via workflow_dispatch.
+# Runs `pre-commit run` for ruff / codespell / gitleaks / EOF / trailing-ws.
+# Called from build.yml on PRs and merge_group; also runnable on demand via
+# workflow_dispatch for manual local-equivalent linting.
 on:
  workflow_call:
  workflow_dispatch:
@@ -12,6 +13,10 @@ permissions:
 jobs:
  pre-commit:
    runs-on: ubuntu-latest
+    env:
+      # Prevents sdist builds → no tar extraction
+      PIP_ONLY_BINARY: ":all:"
+      PIP_DISABLE_PIP_VERSION_CHECK: "1"
    steps:
      - name: Harden Runner
        uses: step-security/harden-runner@ab7a9404c0f3da075243ca237b5fac12c98deaa5 # v2.19.3
@@ -24,13 +29,23 @@ jobs:
          fetch-depth: 0
          persist-credentials: false

-      - name: Install uv
-        uses: astral-sh/setup-uv@fac544c07dec837d0ccb6301d7b5580bf5edae39 # v8.2.0
+      - name: Set up Python
+        uses: actions/setup-python@a309ff8b426b58ec0e2a45f0f869d46889d02405 # v6.2.0
        with:
-          enable-cache: true
+          python-version: 3.12
+          cache: "pip" # caching pip dependencies
+          cache-dependency-path: ./.github/scripts/requirements_pre_commit.txt

-      - name: Install Task
-        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
+      - name: Run Pre-Commit Hooks
+        run: |
+          pip install --require-hashes --only-binary=:all: -r ./.github/scripts/requirements_pre_commit.txt

-      - name: Run pre-commit checks
-        run: task pre-commit
+      - name: Run Pre-Commit
+        run: |
+          pre-commit run ruff --all-files -c .pre-commit-config.yaml
+          pre-commit run ruff-format --all-files -c .pre-commit-config.yaml
+          pre-commit run codespell --all-files -c .pre-commit-config.yaml
+          pre-commit run gitleaks --all-files -c .pre-commit-config.yaml
+          pre-commit run end-of-file-fixer --all-files -c .pre-commit-config.yaml
+          pre-commit run trailing-whitespace --all-files -c .pre-commit-config.yaml
+          git diff --exit-code
@@ -75,7 +75,7 @@ jobs:

      - name: Generate tags for base image
        id: meta
-        uses: docker/metadata-action@80c7e94dd9b9319bd5eb7a0e0fe9291e23a2a2e9 # v6.1.0
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # v6.0.0
        with:
          images: |
            ${{ secrets.DOCKER_HUB_ORG_USERNAME }}/stirling-pdf-base
@@ -78,14 +78,14 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1

      - name: Set up Docker Buildx
        id: buildx
        uses: docker/setup-buildx-action@4d04d5d9486b7bd6fa91e7baf45bbb4f8b9deedd # v4.0.0

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Get version number
        id: versionNumber
        run: echo "versionNumber=$(./gradlew printVersion --quiet | tail -1)" >> $GITHUB_OUTPUT
@@ -129,7 +129,7 @@ jobs:
      - name: Generate tags for latest
        id: meta
        if: env.RUN_MAIN_APP == 'true'
-        uses: docker/metadata-action@80c7e94dd9b9319bd5eb7a0e0fe9291e23a2a2e9 # v6.1.0
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # v6.0.0
        with:
          images: |
            ${{ secrets.DOCKER_HUB_USERNAME }}/s-pdf
@@ -155,9 +155,9 @@ jobs:
          cache-to: type=gha,mode=max,scope=stirling-pdf-latest
          tags: ${{ steps.meta.outputs.tags }}
          labels: ${{ steps.meta.outputs.labels }}
-          # No BASE_VERSION pin: inherit the Dockerfile ARG default (single source of truth).
          build-args: |
            VERSION_TAG=${{ steps.versionNumber.outputs.versionNumber }}
+            BASE_VERSION=1.0.0
          platforms: linux/amd64,linux/arm64/v8
          provenance: true
          sbom: true
@@ -178,7 +178,7 @@ jobs:

      - name: Generate tags for latest-fat
        id: meta-fat
-        uses: docker/metadata-action@80c7e94dd9b9319bd5eb7a0e0fe9291e23a2a2e9 # v6.1.0
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # v6.0.0
        if: env.RUN_MAIN_APP == 'true' && github.ref != 'refs/heads/main' && github.ref != 'refs/heads/testMain'
        with:
          images: |
@@ -222,7 +222,7 @@ jobs:

      - name: Generate tags for ultra-lite
        id: meta-lite
-        uses: docker/metadata-action@80c7e94dd9b9319bd5eb7a0e0fe9291e23a2a2e9 # v6.1.0
+        uses: docker/metadata-action@030e881283bb7a6894de51c315a6bfe6a94e05cf # v6.0.0
        if: env.RUN_MAIN_APP == 'true' && github.ref != 'refs/heads/main' && github.ref != 'refs/heads/testMain'
        with:
          images: |
@@ -22,7 +22,7 @@ jobs:
          egress-policy: audit

      - name: 30 days stale issues
-        uses: actions/stale@eb5cf3af3ac0a1aa4c9c45633dd1ae542a27a899 # v10.3.0
+        uses: actions/stale@b5d41d4e1d5dceea10e7104786b73624c18a190f # v10.2.0
        with:
          repo-token: ${{ secrets.GITHUB_TOKEN }}
          days-before-stale: 30
@@ -48,7 +48,7 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1

      - name: Generate Swagger documentation
        run: ./gradlew :stirling-pdf:generateOpenApiDocs
@@ -63,7 +63,7 @@ jobs:
          SWAGGERHUB_USER: "Frooodle"

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Get version number
        id: versionNumber
        run: echo "versionNumber=$(./gradlew printVersion --quiet | tail -1)" >> $GITHUB_OUTPUT
@@ -58,23 +58,15 @@ jobs:

      - name: Install Python dependencies
        run: |
-          pip install --require-hashes --only-binary=:all: -r ./.github/scripts/requirements_sync_readme.txt
-
-      - name: Install uv
-        uses: astral-sh/setup-uv@fac544c07dec837d0ccb6301d7b5580bf5edae39 # v8.2.0
-        with:
-          enable-cache: true
-
-      - name: Install Task
-        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
+          pip install --require-hashes --only-binary=:all: -r ./.github/scripts/requirements_sync_readme.txt -r ./.github/scripts/requirements_pre_commit.txt

      - name: Sync translation TOML files
        run: |
-          python .github/scripts/check_language_toml.py --reference-file "frontend/editor/public/locales/en-US/translation.toml" --branch main
+          python .github/scripts/check_language_toml.py --reference-file "frontend/editor/public/locales/en-GB/translation.toml" --branch main

-      - name: Sort translation TOML files
+      - name: pre-commit run
        run: |
-          task pre-commit:toml-sort FIX=1
+          pre-commit run toml-sort-fix --all-files

      - name: Commit translation files
        run: |
@@ -108,7 +100,7 @@ jobs:
            This Pull Request was automatically generated to synchronize updates to translation files and documentation. Below are the details of the changes made:

            #### **1. Synchronization of Translation Files**
-            - Updated translation files (`frontend/editor/public/locales/*/translation.toml`) to reflect changes in the reference file `en-US/translation.toml`.
+            - Updated translation files (`frontend/editor/public/locales/*/translation.toml`) to reflect changes in the reference file `en-GB/translation.toml`.
            - Ensured consistency and synchronization across all supported language files.
            - Highlighted any missing or incomplete translations.
            - **Format**: TOML
@@ -16,11 +16,6 @@ on:
        required: false
        type: string
        default: "all"
-      sign:
-        description: "Sign and notarize the bundles."
-        required: false
-        type: boolean
-        default: true
  workflow_dispatch:
    inputs:
      platform:
@@ -33,11 +28,6 @@ on:
          - windows
          - macos
          - linux
-      sign:
-        description: "Sign and notarize the bundles."
-        required: false
-        default: true
-        type: boolean

 permissions:
  contents: read
@@ -125,20 +115,6 @@ jobs:
          toolchain: stable
          targets: ${{ matrix.platform == 'macos-15' && 'aarch64-apple-darwin,x86_64-apple-darwin' || '' }}

-      # Cache the Cargo registry and compiled dependency crates so the build
-      # only recompiles the app crate. Written on main; PRs and the merge queue
-      # restore from it.
-      - name: Cache Rust build
-        uses: Swatinem/rust-cache@c19371144df3bb44fab255c43d04cbc2ab54d1c4 # v2.9.1
-        with:
-          workspaces: frontend/editor/src-tauri
-          # Stable key shared across workflows so the nightly warmer.
-          # rust-cache still appends OS + rustc + Cargo.lock.
-          shared-key: tauri-${{ matrix.name }}
-          save-if: ${{ github.ref == 'refs/heads/main' }}
-          # Save the dependency cache even if a later step fails
-          cache-on-failure: true
-
      - name: Set up x86_64 JDK 25 (macOS universal JRE)
        if: matrix.platform == 'macos-15'
        uses: actions/setup-java@be666c2fcd27ec809703dec50e508c2fdc7f6654 # v5.2.0
@@ -160,10 +136,10 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1

      - name: Setup Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0

      - name: Build universal macOS JRE
        if: matrix.platform == 'macos-15'
@@ -187,7 +163,7 @@ jobs:
      # DigiCert KeyLocker Setup (Cloud HSM)
      - name: Setup DigiCert KeyLocker
        id: digicert-setup
-        if: ${{ inputs.sign && matrix.platform == 'windows-latest' && env.SM_API_KEY != '' && github.ref == 'refs/heads/main' }}
+        if: ${{ matrix.platform == 'windows-latest' && env.SM_API_KEY != '' && github.ref == 'refs/heads/main' }}
        uses: digicert/ssm-code-signing@1d820463733701cf1484c7eb5d7d24a15ca2c454 # v1.2.1
        env:
          SM_API_KEY: ${{ secrets.SM_API_KEY }}
@@ -197,7 +173,7 @@ jobs:
          SM_HOST: ${{ secrets.SM_HOST }}

      - name: Setup DigiCert KeyLocker Certificate
-        if: ${{ inputs.sign && matrix.platform == 'windows-latest' && env.SM_API_KEY != '' && github.ref == 'refs/heads/main' }}
+        if: ${{ matrix.platform == 'windows-latest' && env.SM_API_KEY != '' && github.ref == 'refs/heads/main' }}
        shell: pwsh
        run: |
          Write-Host "Setting up DigiCert KeyLocker environment..."
@@ -232,7 +208,7 @@ jobs:

      # Traditional PFX Certificate Import (fallback if KeyLocker not configured)
      - name: Import Windows Code Signing Certificate
-        if: ${{ inputs.sign && matrix.platform == 'windows-latest' && env.SM_API_KEY == '' && github.ref == 'refs/heads/main' }}
+        if: ${{ matrix.platform == 'windows-latest' && env.SM_API_KEY == '' && github.ref == 'refs/heads/main' }}
        env:
          WINDOWS_CERTIFICATE: ${{ secrets.WINDOWS_CERTIFICATE }}
          WINDOWS_CERTIFICATE_PASSWORD: ${{ secrets.WINDOWS_CERTIFICATE_PASSWORD }}
@@ -263,7 +239,7 @@ jobs:
          }

      - name: Import Apple Developer Certificate
-        if: inputs.sign && matrix.platform == 'macos-15' && env.APPLE_CERTIFICATE != ''
+        if: matrix.platform == 'macos-15' && env.APPLE_CERTIFICATE != ''
        env:
          APPLE_CERTIFICATE: ${{ secrets.APPLE_CERTIFICATE }}
          APPLE_CERTIFICATE_PASSWORD: ${{ secrets.APPLE_CERTIFICATE_PASSWORD }}
@@ -284,7 +260,7 @@ jobs:
          rm certificate.p12

      - name: Verify Certificate
-        if: inputs.sign && matrix.platform == 'macos-15' && env.APPLE_CERTIFICATE != ''
+        if: matrix.platform == 'macos-15' && env.APPLE_CERTIFICATE != ''
        run: |
          echo "Verifying Apple Developer Certificate..."
          KEYCHAIN_PATH=$RUNNER_TEMP/app-signing.keychain-db
@@ -307,7 +283,7 @@ jobs:
          ls -la /usr/bin/hd* || echo "No hd* tools found"

      - name: Preflight smctl
-        if: ${{ inputs.sign && matrix.platform == 'windows-latest' && env.SM_API_KEY != '' && github.ref == 'refs/heads/main' }}
+        if: ${{ matrix.platform == 'windows-latest' && env.SM_API_KEY != '' && github.ref == 'refs/heads/main' }}
        shell: pwsh
        env:
          KEYPAIR_ALIAS: ${{ secrets.SM_KEYPAIR_ALIAS }}
@@ -320,7 +296,7 @@ jobs:
          if ($LASTEXITCODE -ne 0) { Write-Host "[WARN] smctl windows certsync returned non-zero - continuing" }

      - name: Configure Windows code signing
-        if: ${{ inputs.sign && matrix.platform == 'windows-latest' && env.SM_API_KEY != '' && github.ref == 'refs/heads/main' }}
+        if: ${{ matrix.platform == 'windows-latest' && env.SM_API_KEY != '' && github.ref == 'refs/heads/main' }}
        shell: bash
        env:
          KEYPAIR_ALIAS: ${{ secrets.SM_KEYPAIR_ALIAS }}
@@ -339,7 +315,7 @@ jobs:
          EOF

      - name: Import release GPG signing key (Linux)
-        if: inputs.sign && matrix.platform == 'ubuntu-22.04' && env.RELEASE_GPG_PRIVATE_KEY != '' && github.ref == 'refs/heads/main'
+        if: matrix.platform == 'ubuntu-22.04' && env.RELEASE_GPG_PRIVATE_KEY != '' && github.ref == 'refs/heads/main'
        run: |
          echo "$RELEASE_GPG_PRIVATE_KEY" | gpg --batch --import
          gpg --list-secret-keys --keyid-format=long
@@ -356,8 +332,7 @@ jobs:
            exit 1
          fi

-      - name: Build Tauri app (signed)
-        if: inputs.sign
+      - name: Build Tauri app
        uses: tauri-apps/tauri-action@84b9d35b5fc46c1e45415bdb6144030364f7ebc5 # v0.6.2
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
@@ -391,26 +366,6 @@ jobs:
          # failure (#6127 onwards) does not tank deb/rpm uploads.
          args: ${{ matrix.platform == 'ubuntu-22.04' && '--bundles deb,rpm' || matrix.args }}

-      - name: Build Tauri app (unsigned)
-        if: ${{ !inputs.sign }}
-        uses: tauri-apps/tauri-action@84b9d35b5fc46c1e45415bdb6144030364f7ebc5 # v0.6.2
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          SIGN: "0"
-          TAURI_SIGNING_PRIVATE_KEY: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY }}
-          TAURI_SIGNING_PRIVATE_KEY_PASSWORD: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY_PASSWORD }}
-          VITE_SUPABASE_PUBLISHABLE_DEFAULT_KEY: ${{ secrets.VITE_SUPABASE_PUBLISHABLE_DEFAULT_KEY || 'sb_publishable_UHz2SVRF5mvdrPHWkRteyA_yNlZTkYb' }}
-          VITE_SAAS_SERVER_URL: ${{ secrets.VITE_SAAS_SERVER_URL || 'https://app.stirlingpdf.com' }}
-          VITE_SAAS_BACKEND_API_URL: ${{ secrets.VITE_SAAS_BACKEND_API_URL || 'https://api.stirlingpdf.com' }}
-          CI: true
-        with:
-          projectPath: ./frontend/editor
-          tauriScript: npx tauri
-          # Linux: build deb+rpm only here. AppImage runs in its own
-          # continue-on-error step below so its persistent linuxdeploy
-          # failure (#6127 onwards) does not tank deb/rpm uploads.
-          args: ${{ matrix.platform == 'ubuntu-22.04' && '--bundles deb,rpm' || matrix.args }}
-
      # AppImage is decoupled so its linuxdeploy run gets a fresh process
      # (rpm scratch state torn down) and its failure can't tank deb/rpm.
      - name: Build Tauri app (Linux AppImage)
@@ -419,7 +374,7 @@ jobs:
        uses: tauri-apps/tauri-action@84b9d35b5fc46c1e45415bdb6144030364f7ebc5 # v0.6.2
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-          SIGN: ${{ (inputs.sign && env.RELEASE_GPG_PRIVATE_KEY != '' && github.ref == 'refs/heads/main') && '1' || '0' }}
+          SIGN: ${{ (env.RELEASE_GPG_PRIVATE_KEY != '' && github.ref == 'refs/heads/main') && '1' || '0' }}
          APPIMAGETOOL_SIGN_PASSPHRASE: ${{ secrets.RELEASE_GPG_PASSPHRASE }}
          SIGN_KEY: ${{ vars.RELEASE_GPG_FINGERPRINT }}
          TAURI_SIGNING_PRIVATE_KEY: ${{ secrets.TAURI_SIGNING_PRIVATE_KEY }}
@@ -434,7 +389,7 @@ jobs:
          args: --bundles appimage

      - name: Clear release GPG key from runner keyring (Linux)
-        if: always() && inputs.sign && matrix.platform == 'ubuntu-22.04' && env.RELEASE_GPG_PRIVATE_KEY != '' && github.ref == 'refs/heads/main'
+        if: always() && matrix.platform == 'ubuntu-22.04' && env.RELEASE_GPG_PRIVATE_KEY != '' && github.ref == 'refs/heads/main'
        env:
          RELEASE_GPG_FINGERPRINT: ${{ vars.RELEASE_GPG_FINGERPRINT }}
        run: |
@@ -444,7 +399,7 @@ jobs:
          fi

      - name: Verify notarization (macOS only)
-        if: inputs.sign && matrix.platform == 'macos-15'
+        if: matrix.platform == 'macos-15'
        run: |
          echo "🔍 Verifying notarization status..."
          cd ./frontend/editor/src-tauri/target
@@ -482,7 +437,7 @@ jobs:
      # Verify the MSI AND the inner exe extracted from it are signed.
      # The inner exe is what gets installed on users' machines and what AV scans.
      - name: Verify Windows Code Signature
-        if: inputs.sign && matrix.platform == 'windows-latest' && env.SM_API_KEY != '' && github.ref == 'refs/heads/main'
+        if: matrix.platform == 'windows-latest' && env.SM_API_KEY != '' && github.ref == 'refs/heads/main'
        shell: pwsh
        run: |
          $allSigned = $true
@@ -106,11 +106,11 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1
          cache-disabled: true

      - name: Install Task
-        uses: go-task/setup-task@01a4adf9db2d14c1de7a560f09170b6e0df736aa # v2.1.0
+        uses: go-task/setup-task@3be4020d41929789a01026e0e427a4321ce0ad44 # v2.0.0
      - name: Build application
        run: task backend:build
        env:
@@ -155,22 +155,9 @@ jobs:
            echo "platforms=linux/amd64,linux/arm64/v8" >> "$GITHUB_OUTPUT"
          fi

-      # Base-changed PRs build the embedded image with the local docker driver
-      # so the locally-built stirling-pdf-base:pr-test (in the daemon image
-      # store) resolves. A buildx container builder cannot see it and would try
-      # to pull it from a registry, which fails. Single-platform, no gha cache.
-      - name: Build ${{ matrix.docker-rev }} against local base (PR base change)
-        if: github.event_name == 'pull_request' && inputs.docker-base-changed == 'true'
-        run: |
-          DOCKER_BUILDKIT=1 docker build \
-            --build-arg BASE_IMAGE=${{ steps.build-params.outputs.base_image }} \
-            --file ./${{ matrix.docker-rev }} \
-            --tag stirling-pdf-embedded:pr-test \
-            .
-
      - name: Build ${{ matrix.docker-rev }} (Depot)
        if: env.USE_DEPOT == 'true'
-        uses: depot/build-push-action@98e78adca7817480b8185f474a400b451d74e287 # v1.16.0
+        uses: depot/build-push-action@5f3b3c2e5a00f0093de47f657aeaefcedff27d18 # v1.16.0
        with:
          project: ${{ vars.DEPOT_PROJECT_ID }}
          context: .
@@ -182,10 +169,8 @@ jobs:
          provenance: true
          sbom: true

-      # Fork PRs that did NOT change the base use the buildx container builder
-      # (multi-platform + gha cache) against the published base image.
      - name: Build ${{ matrix.docker-rev }} (Docker fork fallback)
-        if: env.USE_DEPOT != 'true' && inputs.docker-base-changed != 'true'
+        if: env.USE_DEPOT != 'true'
        uses: docker/build-push-action@bcafcacb16a39f128d818304e6c9c0c18556b85f # v7.1.0
        with:
          builder: ${{ steps.buildx.outputs.name }}
@@ -245,7 +230,7 @@ jobs:

      - name: Build docker/unoserver/Dockerfile (Depot)
        if: env.USE_DEPOT == 'true'
-        uses: depot/build-push-action@98e78adca7817480b8185f474a400b451d74e287 # v1.16.0
+        uses: depot/build-push-action@5f3b3c2e5a00f0093de47f657aeaefcedff27d18 # v1.16.0
        with:
          project: ${{ vars.DEPOT_PROJECT_ID }}
          context: .
@@ -51,7 +51,7 @@ jobs:
      - name: Setup Gradle
        uses: gradle/actions/setup-gradle@50e97c2cd7a37755bbfafc9c5b7cafaece252f6e # v6.1.0
        with:
-          gradle-version: 9.6.0
+          gradle-version: 9.3.1

      - name: Build with Gradle
        run: ./gradlew build
@@ -83,7 +83,7 @@ jobs:

      - name: Build and push test image (Depot)
        if: env.USE_DEPOT == 'true'
-        uses: depot/build-push-action@98e78adca7817480b8185f474a400b451d74e287 # v1.16.0
+        uses: depot/build-push-action@5f3b3c2e5a00f0093de47f657aeaefcedff27d18 # v1.16.0
        with:
          project: ${{ vars.DEPOT_PROJECT_ID }}
          context: .
@@ -129,7 +129,7 @@ jobs:
              environment:
                DISABLE_ADDITIONAL_FEATURES: "true"
                SECURITY_ENABLELOGIN: "false"
-                SYSTEM_DEFAULTLOCALE: en-US
+                SYSTEM_DEFAULTLOCALE: en-GB
                UI_APPNAME: "Stirling-PDF Test"
                UI_HOMEDESCRIPTION: "Test Deployment"
                UI_APPNAMENAVBAR: "Test"
@@ -46,12 +46,6 @@ app/core/storage/
 # These are generated by npm build and should not be committed
 app/core/src/main/resources/static/assets/
 app/core/src/main/resources/static/index.html
-# Prerendered per-route SPA pages (OG/social-preview), e.g. compress.html. api-landing.html is source.
-app/core/src/main/resources/static/*.html
-!app/core/src/main/resources/static/api-landing.html
-!app/core/src/main/resources/static/mobile-upload.html
-# Prerendered nested-route pages (e.g. settings/people.html)
-app/core/src/main/resources/static/settings/
 app/core/src/main/resources/static/locales/
 app/core/src/main/resources/static/Login/
 app/core/src/main/resources/static/classic-logo/
@@ -59,8 +53,6 @@ app/core/src/main/resources/static/modern-logo/
 app/core/src/main/resources/static/og_images/
 app/core/src/main/resources/static/samples/
 app/core/src/main/resources/static/manifest-classic.json
-app/core/src/main/resources/static/og-metadata.json
-app/core/src/main/resources/static/sw-folder-retry.js
 app/core/src/main/resources/static/robots.txt
 app/core/src/main/resources/static/pdfium/
 app/core/src/main/resources/static/pdfjs/
@@ -1,29 +1,5 @@
-# PostHog project-level key - phc_ prefix keys are public/client-side by design
+# PostHog project-level key — phc_ prefix keys are public/client-side by design
 # (PostHog client-side tracking embeds them in the browser bundle). Committed
 # intentionally in #6150 so engine/.env has a working default, with real
 # credentials overridden via engine/.env.local.
 engine/.env:generic-api-key:41
-
-# MCP test fixtures / harness - no real secrets:
-# - test-only API key constant in an integration test
-# - JDBC URL + throwaway Keycloak creds in the local test compose
-# - placeholder / shell-variable Bearer headers in curl-based validation scripts
-app/proprietary/src/test/java/stirling/software/proprietary/mcp/security/McpApiKeyIntegrationTest.java:generic-api-key:40
-testing/compose/docker-compose-keycloak-mcp.yml:generic-api-key:25
-testing/compose/validate-mcp-apikey.sh:curl-auth-header:73
-testing/compose/validate-mcp-test.sh:curl-auth-header:92
-testing/compose/validate-mcp-test.sh:curl-auth-header:116
-
-# Storybook example showing curl with a fake Bearer token placeholder (sk_live_a3f8...).
-frontend/editor/src/proprietary/ui/CodeBlock.stories.tsx:curl-auth-header:5
-
-# Truncated placeholder API key in portal docs example (sk_live_8f2c...e10) - not a real secret.
-frontend/editor/src/portal/components/docs/GettingStartedSection.tsx:generic-api-key:30
-
-# False positive: generic-api-key matches the Java type name "X509Certificate"
-# in a method signature (CreateSignatureBase.resolveSignatureAlgorithm) - not a secret.
-app/core/src/main/java/org/apache/pdfbox/examples/signature/CreateSignatureBase.java:generic-api-key:224
-
-# Supabase publishable key (public by design, RLS-protected) used as a CI fallback
-# default in the tauri-build workflow when the GitHub secret is unset - not a real secret.
-.github/workflows/tauri-build.yml:generic-api-key:402
@@ -1,14 +1,52 @@
-# The actual checks live in .taskfiles/pre-commit.yml (with helper scripts under
-# scripts/pre-commit/) and are driven by Task. This hook just delegates to `task
-# pre-commit` so the git pre-commit hook, CI and a manual `task pre-commit` all
-# run the exact same thing. Requires `task` and `uv` on PATH. To auto-fix instead
-# of only checking, run `task pre-commit:fix`.
 repos:
-  - repo: local
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.15.14
    hooks:
-      - id: task-pre-commit
-        name: task pre-commit
-        entry: task pre-commit
-        language: system
-        pass_filenames: false
-        always_run: true
+      - id: ruff
+        args:
+          - --fix
+          - --line-length=127
+        files: ^((\.github/scripts|scripts|app/core/src/main/resources/static/python)/.+)?[^/]+\.py$
+        exclude: (split_photos.py)
+      - id: ruff-format
+        files: ^((\.github/scripts|scripts|app/core/src/main/resources/static/python)/.+)?[^/]+\.py$
+        exclude: (split_photos.py)
+  - repo: https://github.com/codespell-project/codespell
+    rev: v2.4.2
+    hooks:
+      - id: codespell
+        args:
+          - --ignore-words-list=thirdParty,tabEl,tabEls,Sie,ist,fulfilment
+          - --skip="./.*,*.csv,*.json,*.ambr"
+          - --quiet-level=2
+        files: \.(html|css|js|py|md)$
+        exclude: (.vscode|.devcontainer|app/core/src/main/resources|app/proprietary/src/main/resources|frontend/editor/public/vendor|Dockerfile|.*/pdfjs.*|.*/thirdParty.*|bootstrap.*|.*\.min\..*|.*diff\.js)
+  - repo: https://github.com/gitleaks/gitleaks
+    rev: v8.30.0
+    hooks:
+      - id: gitleaks
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v6.0.0
+    hooks:
+      - id: end-of-file-fixer
+        files: ^.*(\.js|\.java|\.py|\.yml)$
+        exclude: ^(.*/pdfjs.*|.*/thirdParty.*|bootstrap.*|.*\.min\..*|.*diff\.js|\.github/workflows/.*$)
+      - id: trailing-whitespace
+        files: ^.*(\.js|\.java|\.py|\.yml)$
+        exclude: ^(.*/pdfjs.*|.*/thirdParty.*|bootstrap.*|.*\.min\..*|.*diff\.js|\.github/workflows/.*$)
+  - repo: https://github.com/pappasam/toml-sort
+    rev: v0.24.4
+    hooks:
+      - id: toml-sort-fix
+        files: frontend/editor/public/locales/.*\.toml$
+        args: ['--in-place', '--all', '--ignore-case']
+  # - repo: https://github.com/thibaudcolas/pre-commit-stylelint
+  #   rev: v16.21.1
+  #   hooks:
+  #     - id: stylelint
+  #       additional_dependencies:
+  #         - stylelint@16.21.1
+  #         - stylelint-config-standard@38.0.0
+  #         - "@stylistic/stylelint-plugin@3.1.3"
+  #       files: \.(css)$
+  #       args: [--fix]
@@ -18,36 +18,17 @@ version: '3'
 tasks:
  dev:
    desc: "Start backend dev server"
-    cmds:
-      - task: dev:proprietary
-        vars:
-          PORT: '{{.PORT}}'
-          AIENGINE_URL: '{{.AIENGINE_URL}}'
-          AIENGINE_ENABLED: '{{.AIENGINE_ENABLED}}'
-          AIENGINE_TIMEOUTSECONDS: '{{.AIENGINE_TIMEOUTSECONDS}}'
-          SECURITY_ENABLELOGIN: '{{.SECURITY_ENABLELOGIN}}'
-          POLICIES_ENABLED: '{{.POLICIES_ENABLED}}'
-
-  dev:proprietary:
-    desc: "Start backend dev server in proprietary mode"
-    # `dotenv:` reads from the root Taskfile's directory (".") because this
-    # subtaskfile is included with `dir: .`. Local overrides in
-    # .env.proprietary.local win over the committed .env.proprietary defaults.
-    dotenv: ['app/.env.proprietary.local', 'app/.env.proprietary']
    ignore_error: true
    vars:
      PORT: '{{.PORT | default "8080"}}'
      AIENGINE_URL: '{{.AIENGINE_URL | default ""}}'
-      AIENGINE_ENABLED: '{{.AIENGINE_ENABLED | default "false"}}'
      AIENGINE_TIMEOUTSECONDS: '{{.AIENGINE_TIMEOUTSECONDS | default "120"}}'
-      SECURITY_ENABLELOGIN: '{{.SECURITY_ENABLELOGIN | default ""}}'
-      POLICIES_ENABLED: '{{.POLICIES_ENABLED | default ""}}'
    env:
      SERVER_PORT: '{{.PORT}}'
    cmds:
-      - cmd: '{{if .AIENGINE_URL}}AIENGINE_URL={{.AIENGINE_URL}} AIENGINE_ENABLED={{.AIENGINE_ENABLED}} AIENGINE_TIMEOUTSECONDS={{.AIENGINE_TIMEOUTSECONDS}} {{end}}{{if .SECURITY_ENABLELOGIN}}SECURITY_ENABLELOGIN={{.SECURITY_ENABLELOGIN}} {{end}}{{if .POLICIES_ENABLED}}POLICIES_ENABLED={{.POLICIES_ENABLED}} {{end}}cmd /c ".\gradlew.bat :stirling-pdf:bootRun"'
+      - cmd: '{{if .AIENGINE_URL}}AIENGINE_URL={{.AIENGINE_URL}} AIENGINE_ENABLED=true AIENGINE_TIMEOUTSECONDS={{.AIENGINE_TIMEOUTSECONDS}} {{end}}cmd /c ".\gradlew.bat :stirling-pdf:bootRun"'
        platforms: [windows]
-      - cmd: '{{if .AIENGINE_URL}}AIENGINE_URL={{.AIENGINE_URL}} AIENGINE_ENABLED={{.AIENGINE_ENABLED}} AIENGINE_TIMEOUTSECONDS={{.AIENGINE_TIMEOUTSECONDS}} {{end}}{{if .SECURITY_ENABLELOGIN}}SECURITY_ENABLELOGIN={{.SECURITY_ENABLELOGIN}} {{end}}{{if .POLICIES_ENABLED}}POLICIES_ENABLED={{.POLICIES_ENABLED}} {{end}}./gradlew :stirling-pdf:bootRun'
+      - cmd: '{{if .AIENGINE_URL}}AIENGINE_URL={{.AIENGINE_URL}} AIENGINE_ENABLED=true AIENGINE_TIMEOUTSECONDS={{.AIENGINE_TIMEOUTSECONDS}} {{end}}./gradlew :stirling-pdf:bootRun'
        platforms: [linux, darwin]

  dev:bundled:
@@ -69,15 +50,9 @@ tasks:
      PORT: '{{.PORT | default "8080"}}'
      # Override to "" to run the pure `saas` profile against your own SAAS_DB_*.
      PROFILES: '{{.PROFILES | default "dev"}}'
-      AIENGINE_URL: '{{.AIENGINE_URL | default ""}}'
-      AIENGINE_ENABLED: '{{.AIENGINE_ENABLED | default "false"}}'
-      AIENGINE_TIMEOUTSECONDS: '{{.AIENGINE_TIMEOUTSECONDS | default "120"}}'
    env:
      SERVER_PORT: '{{.PORT}}'
      STIRLING_FLAVOR: saas
-      AIENGINE_URL: '{{.AIENGINE_URL}}'
-      AIENGINE_ENABLED: '{{.AIENGINE_ENABLED}}'
-      AIENGINE_TIMEOUTSECONDS: '{{.AIENGINE_TIMEOUTSECONDS}}'
    cmds:
      - cmd: cmd /c ".\gradlew.bat :stirling-pdf:bootRun {{if .PROFILES}}--args=\"--spring.profiles.include={{.PROFILES}}\"{{end}}"
        platforms: [windows]
@@ -1,14 +1,7 @@
 version: '3'

 vars:
-  # jdk.dynalink is required by VeraPDF (PDF/A validation); without it the bundled JRE throws
-  # NoClassDefFoundError: jdk/dynalink/Namespace at runtime in get-info-on-pdf and verify-pdf
-  JLINK_MODULES: "java.base,java.compiler,java.desktop,java.instrument,java.logging,java.management,java.naming,java.net.http,java.prefs,java.rmi,java.scripting,java.security.jgss,java.security.sasl,java.sql,java.transaction.xa,java.xml,java.xml.crypto,jdk.crypto.ec,jdk.crypto.cryptoki,jdk.unsupported,jdk.dynalink"
-
-  # Minimum Java major the bundled JRE must be. Keep in sync with build.gradle
-  # `modernJavaVersion` - the app JAR is compiled for this, so an older runtime
-  # fails at launch with UnsupportedClassVersionError. Enforced by jlink:verify.
-  REQUIRED_JAVA: "25"
+  JLINK_MODULES: "java.base,java.compiler,java.desktop,java.instrument,java.logging,java.management,java.naming,java.net.http,java.prefs,java.rmi,java.scripting,java.security.jgss,java.security.sasl,java.sql,java.transaction.xa,java.xml,java.xml.crypto,jdk.crypto.ec,jdk.crypto.cryptoki,jdk.unsupported"

  # Override via JPDFIUM_PLATFORMS env (csv of platform keys, or 'all').
  JPDFIUM_PLATFORMS:
@@ -69,21 +62,21 @@ tasks:
    deps: [prepare]
    dir: editor
    cmds:
-      - npx tauri build --bundles app --config '{"bundle":{"createUpdaterArtifacts":false}}'
+      - npx tauri build --bundles app

  build:dev:windows:
    desc: "Build Tauri desktop NSIS installer (Windows)"
    deps: [prepare]
    dir: editor
    cmds:
-      - npx tauri build --bundles nsis --config '{"bundle":{"createUpdaterArtifacts":false}}'
+      - npx tauri build --bundles nsis

  build:dev:linux:
    desc: "Build Tauri desktop AppImage (Linux)"
    deps: [prepare]
    dir: editor
    cmds:
-      - npx tauri build --bundles appimage --config '{"bundle":{"createUpdaterArtifacts":false}}'
+      - npx tauri build --bundles appimage

  test:
    desc: "Run Tauri/Cargo tests"
@@ -107,20 +100,6 @@ tasks:
  jlink:
    desc: "Build backend JAR and create JLink runtime for Tauri"
    deps: [jlink:jar, jlink:runtime]
-    # Runs after the runtime is in place. Lives here (not in jlink:runtime's
-    # cmds) so it still fires when jlink:runtime short-circuits on its `status:`
-    # check and reuses an existing runtime/jre - that reuse path is exactly how
-    # a stale, too-old JRE slips through.
-    cmds:
-      - task: jlink:verify
-
-  jlink:verify:
-    desc: "Fail the build if the bundled JRE is older than the app JAR requires"
-    dir: editor
-    env:
-      REQUIRED_JAVA: "{{.REQUIRED_JAVA}}"
-    cmds:
-      - node scripts/verify-bundled-jre.mjs src-tauri/runtime/jre/release

  jlink:jar:
    desc: "Build backend JAR for Tauri bundling (host-OS natives only by default)"
@@ -146,35 +125,14 @@ tasks:
    cmds:
      - rm -rf runtime/jre
      - mkdir -p runtime
-      # Pin jlink to JAVA_HOME so the bundled JRE matches the JDK the build
-      # uses. Bare `jlink` on PATH can resolve to an older system Java (the
-      # ubuntu runner ships Java 11), producing a runtime jlink:verify rejects.
-      #
-      # jdk.crypto.mscapi (the Windows certificate store / SunMSCAPI provider, used by
-      # hardware-backed cert signing) is a Windows-only module - it only exists in a Windows
-      # JDK's jmods, so it is added on Windows only or jlink fails to resolve it elsewhere.
-      - cmd: |
-          JLINK="${JAVA_HOME:+$JAVA_HOME/bin/}jlink"
-          JLINK_COMPRESS="$("$JLINK" --help 2>&1 | grep -q 'zip-\[0-9\]' && echo zip-6 || echo 2)"
-          "$JLINK" \
-            --add-modules {{.JLINK_MODULES}},jdk.crypto.mscapi \
-            --strip-debug \
-            --compress="$JLINK_COMPRESS" \
-            --no-header-files \
-            --no-man-pages \
-            --output runtime/jre
-        platforms: [windows]
-      - cmd: |
-          JLINK="${JAVA_HOME:+$JAVA_HOME/bin/}jlink"
-          JLINK_COMPRESS="$("$JLINK" --help 2>&1 | grep -q 'zip-\[0-9\]' && echo zip-6 || echo 2)"
-          "$JLINK" \
-            --add-modules {{.JLINK_MODULES}} \
-            --strip-debug \
-            --compress="$JLINK_COMPRESS" \
-            --no-header-files \
-            --no-man-pages \
-            --output runtime/jre
-        platforms: [linux, darwin]
+      - >-
+        jlink
+        --add-modules {{.JLINK_MODULES}}
+        --strip-debug
+        --compress=zip-6
+        --no-header-files
+        --no-man-pages
+        --output runtime/jre
      # jlink emits its files mode 444 (read-only). Tauri's build-script
      # resource copier preserves source permissions when staging
      # `runtime/jre/**/*` into `target/<profile>/runtime/jre/...`, so the
@@ -20,11 +20,6 @@ tasks:
    cmds:
      - docker build -t stirling-pdf-ultra-lite -f {{.EMBEDDED_DIR}}/Dockerfile.ultra-lite .

-  build:backend:
-    desc: "Build backend-only Docker image (no embedded frontend)"
-    cmds:
-      - docker build -t stirling-pdf-backend -f docker/backend/Dockerfile .
-
  build:frontend:
    desc: "Build frontend-only Docker image"
    cmds:
@@ -212,55 +212,3 @@ tasks:
    desc: "Stop the SAML keycloak test environment"
    cmds:
      - docker compose -f testing/compose/docker-compose-keycloak-saml.yml down -v
-
-  mcp:up:
-    desc: "Start the MCP keycloak test environment (Stirling as OAuth resource server)"
-    summary: |
-      Brings up Keycloak (OAuth authorization server) + Stirling configured as an
-      MCP resource server, then you can exercise /mcp with real Keycloak tokens.
-      Set LICENSE_KEY=<KEY> to skip the interactive license prompt:
-        task e2e:mcp:up LICENSE_KEY=abc123
-      Pass extra flags via -- :
-        task e2e:mcp:up -- --validate --nobuild
-    ignore_error: true
-    cmds:
-      - bash testing/compose/start-mcp-test.sh {{if .LICENSE_KEY}}--license-key "{{.LICENSE_KEY}}"{{end}} {{.CLI_ARGS}}
-
-  mcp:manual:
-    desc: "Start the MCP keycloak test env in manual mode (prints URLs + a live token for your client)"
-    summary: |
-      Brings the stack up and prints copy-paste URLs/commands plus a freshly minted
-      access token so you can drive your own MCP client (Inspector, curl, ...).
-        task e2e:mcp:manual LICENSE_KEY=<your-license-key>
-      Add --nobuild if the images are already built:
-        task e2e:mcp:manual LICENSE_KEY=<your-license-key> -- --nobuild
-    ignore_error: true
-    cmds:
-      - bash testing/compose/start-mcp-test.sh --manual {{if .LICENSE_KEY}}--license-key "{{.LICENSE_KEY}}"{{end}} {{.CLI_ARGS}}
-
-  mcp:apikey:
-    desc: "Start the MCP test env in API-KEY manual mode (no OAuth/IdP): mints a key + prints client settings"
-    summary: |
-      Brings Stirling up in apikey auth mode and prints copy-paste client settings with a freshly
-      minted X-API-KEY - ideal for clients whose OAuth layer can't reach localhost.
-        task e2e:mcp:apikey LICENSE_KEY=<your-license-key>
-      Add --nobuild if images are already built:
-        task e2e:mcp:apikey LICENSE_KEY=<your-license-key> -- --nobuild
-    ignore_error: true
-    cmds:
-      - bash testing/compose/start-mcp-test.sh --apikey {{if .LICENSE_KEY}}--license-key "{{.LICENSE_KEY}}"{{end}} {{.CLI_ARGS}}
-
-  mcp:validate:
-    desc: "Validate the running MCP keycloak test environment end-to-end (oauth mode + real MCP SDK client)"
-    cmds:
-      - bash testing/compose/validate-mcp-test.sh
-
-  mcp:validate-apikey:
-    desc: "Validate the MCP server in API-KEY auth mode (mints a key + real MCP SDK client), then restore oauth"
-    cmds:
-      - bash testing/compose/validate-mcp-apikey.sh
-
-  mcp:down:
-    desc: "Stop the MCP keycloak test environment"
-    cmds:
-      - docker compose -f testing/compose/docker-compose-keycloak-mcp.yml down -v
@@ -33,7 +33,7 @@ tasks:
    env:
      PYTHONUNBUFFERED: "1"
    cmds:
-      - uv run uvicorn stirling.api.app:app --host 0.0.0.0 --port {{.PORT}} --workers "${STIRLING_ENGINE_WORKERS:-4}"
+      - uv run uvicorn stirling.api.app:app --host 0.0.0.0 --port {{.PORT}}

  dev:
    desc: "Start engine dev server with hot reload"
@@ -40,21 +40,6 @@ tasks:
    cmds:
      - node editor/scripts/generate-icons.js

-  prepare:og:
-    internal: true
-    run: when_changed
-    desc: "Regenerate OG/social-preview metadata from the tool registry"
-    cmds:
-      - node editor/scripts/generate-og-metadata.mjs
-    sources:
-      - editor/src/core/types/toolId.ts
-      - editor/src/core/utils/urlMapping.ts
-      - editor/src/core/data/useTranslatedToolRegistry.tsx
-      - editor/public/og_images/*.png
-    generates:
-      - editor/src/core/data/ogImageMap.json
-      - editor/public/og-metadata.json
-
  prepare:
    desc: "Set up dev environment"
    run: when_changed
@@ -64,7 +49,6 @@ tasks:
      - task: prepare:env
        vars: { MODE: '{{.MODE}}' }
      - prepare:icons
-      - prepare:og

  # ============================================================
  # Development
@@ -80,12 +64,6 @@ tasks:
      OPEN: '{{.OPEN | default ""}}'
    env:
      BACKEND_URL: '{{.BACKEND_URL}}'
-      # Dev-only browser-tab label so concurrent worktrees are distinguishable.
-      # Only the worktree folder basename (e.g. "wt1") is exposed — never the
-      # full path, hostname, or user. Consumed at dev-serve time by vite.config
-      # and dropped from production builds.
-      STIRLING_DEV_LABEL:
-        sh: basename "$(git rev-parse --show-toplevel 2>/dev/null || pwd)"
    cmds:
      - npx vite editor --mode {{.MODE}} --port {{.PORT}}{{if .OPEN}} --open{{end}}

@@ -134,6 +112,12 @@ tasks:
      - task: dev:_run
        vars: { MODE: prototypes, PORT: '{{.PORT}}', BACKEND_URL: '{{.BACKEND_URL}}', OPEN: '{{.OPEN}}' }

+  dev:portal:
+    desc: "Start developer portal dev server"
+    deps: [install]
+    cmds:
+      - npx vite portal --port {{.PORT | default "5173"}}{{if .OPEN}} --open{{end}}
+
  # ============================================================
  # Build
  # ============================================================
@@ -153,10 +137,8 @@ tasks:
  build:proprietary:
    desc: "Build for proprietary mode"
    deps: [prepare]
-    vars:
-      PREVIEW: '{{.PREVIEW | default ""}}'
    cmds:
-      - '{{if .PREVIEW}}VITE_BUILD_FOR_PREVIEW=1 {{end}}npx vite build editor --mode proprietary'
+      - npx vite build editor --mode proprietary

  build:saas:
    desc: "Build for SaaS mode"
@@ -180,6 +162,11 @@ tasks:
    cmds:
      - npx vite build editor --mode prototypes

+  build:portal:
+    desc: "Build developer portal"
+    deps: [install]
+    cmds:
+      - npx vite build portal

  storybook:
    desc: "Start Storybook dev server"
@@ -215,8 +202,8 @@ tasks:
    deps: [install]
    cmds:
      # Globs so dpdm walks the whole tree. dpdm expands the braces itself, so this is
-      # shell-agnostic. Covers the whole editor tree, including the portal layer.
-      - npx dpdm "editor/src/**/*.{ts,tsx}" --circular --no-warning --no-tree --exit-code circular:1
+      # shell-agnostic. Covers editor, portal, and the shared design system.
+      - npx dpdm "editor/src/**/*.{ts,tsx}" "portal/src/**/*.{ts,tsx}" "shared/**/*.{ts,tsx}" --circular --no-warning --no-tree --exit-code circular:1

  lint:fix:
    desc: "Auto-fix lint issues"
@@ -247,26 +234,17 @@ tasks:
    cmds:
      - task: typecheck:proprietary

-  typecheck:_run:
-    internal: true
-    env:
-      CI: '{{ .CI | default "false" }}'
-    cmds:
-      - '{{ if eq .CI "true" }}npx tsc{{ else }}npx tsgo{{ end }} --noEmit --project {{.PROJECT}}'
-
  typecheck:core:
    desc: "Typecheck core build variant"
    deps: [prepare]
    cmds:
-      - task: typecheck:_run
-        vars: { PROJECT: editor/src/core/tsconfig.json }
+      - npx tsc --noEmit --project editor/src/core/tsconfig.json

  typecheck:proprietary:
    desc: "Typecheck proprietary build variant"
    deps: [prepare]
    cmds:
-      - task: typecheck:_run
-        vars: { PROJECT: editor/src/proprietary/tsconfig.json }
+      - npx tsc --noEmit --project editor/src/proprietary/tsconfig.json

  typecheck:saas:
    desc: "Typecheck SaaS build variant"
@@ -274,8 +252,7 @@ tasks:
      - task: prepare
        vars: { MODE: saas }
    cmds:
-      - task: typecheck:_run
-        vars: { PROJECT: editor/src/saas/tsconfig.json }
+      - npx tsc --noEmit --project editor/src/saas/tsconfig.json

  typecheck:desktop:
    desc: "Typecheck desktop build variant"
@@ -283,36 +260,31 @@ tasks:
      - task: prepare
        vars: { MODE: desktop }
    cmds:
-      - task: typecheck:_run
-        vars: { PROJECT: editor/src/desktop/tsconfig.json }
-
-  typecheck:cloud:
-    desc: "Typecheck cloud shared layer (standalone)"
-    deps: [prepare]
-    cmds:
-      - task: typecheck:_run
-        vars: { PROJECT: editor/src/cloud/tsconfig.json }
+      - npx tsc --noEmit --project editor/src/desktop/tsconfig.json

  typecheck:scripts:
    desc: "Typecheck scripts"
    deps: [prepare]
    cmds:
-      - task: typecheck:_run
-        vars: { PROJECT: editor/scripts/tsconfig.json }
+      - npx tsc --noEmit --project editor/scripts/tsconfig.json

  typecheck:prototypes:
    desc: "Typecheck prototypes build variant"
    deps: [prepare]
    cmds:
-      - task: typecheck:_run
-        vars: { PROJECT: editor/src/prototypes/tsconfig.json }
+      - npx tsc --noEmit --project editor/src/prototypes/tsconfig.json

  typecheck:portal:
    desc: "Typecheck developer portal build variant"
    deps: [install]
    cmds:
-      - task: typecheck:_run
-        vars: { PROJECT: editor/src/portal/tsconfig.json }
+      - npx tsc --noEmit --project portal/tsconfig.json
+
+  typecheck:shared:
+    desc: "Typecheck the shared design system"
+    deps: [install]
+    cmds:
+      - npx tsc --noEmit --project shared/tsconfig.json

  typecheck:all:
    desc: "Typecheck all build variants"
@@ -321,10 +293,10 @@ tasks:
      - task: typecheck:proprietary
      - task: typecheck:saas
      - task: typecheck:desktop
-      - task: typecheck:cloud
      - task: typecheck:scripts
      - task: typecheck:prototypes
      - task: typecheck:portal
+      - task: typecheck:shared

  # ============================================================
  # Quality Gate
@@ -338,21 +310,14 @@ tasks:
      - task: format:check
      - task: test

-  og:check:
-    desc: "Fail if committed OG/social-preview metadata is out of date"
-    cmds:
-      - node editor/scripts/generate-og-metadata.mjs --check
-
  check:all:
    desc: "Full CI quality gate"
    cmds:
-      # Runs first, before prepare regenerates: guards the committed og-metadata.json /
-      # ogImageMap.json that the Cloudflare Pages (plain `vite build`) deploy relies on.
-      - task: og:check
      - task: typecheck:all
      - task: lint
      - task: format:check
      - task: build
+      - task: build:portal
      - task: test
      - task: storybook:build

@@ -362,24 +327,19 @@ tasks:

  test:
    desc: "Run tests"
-    cmds:
-      - task: test:editor
-
-  test:editor:
-    desc: "Run editor tests"
-    deps: [prepare]
+    deps: [install]
    cmds:
      - npx vitest run --root editor

  test:watch:
    desc: "Run tests in watch mode"
-    deps: [prepare]
+    deps: [install]
    cmds:
      - npx vitest --watch --root editor

  test:coverage:
    desc: "Run tests with coverage (one-shot; CI-friendly)."
-    deps: [prepare]
+    deps: [install]
    cmds:
      # `vitest run` makes this CI-safe (the bare `vitest` form enters watch
      # mode). Explicit reporter list because v8 + json-summary is what the
@@ -402,37 +362,8 @@ tasks:
  # Code Generation
  # ============================================================

-  tool-models:
-    desc: "Generate tool API types from the Java OpenAPI spec"
-    deps: [install, ":backend:swagger"]
-    cmds:
-      - npx tsx editor/scripts/generate-tool-api-types.mts --spec ../SwaggerDoc.json --output editor/src/core/types/toolApiTypes.ts
-    sources:
-      - editor/scripts/generate-tool-api-types.mts
-      - ../SwaggerDoc.json
-    generates:
-      - editor/src/core/types/toolApiTypes.ts
-
-  tool-models:check:
-    desc: "Fail if committed tool API types are out of date"
-    deps: [install, ":backend:swagger"]
-    cmds:
-      - npx tsx editor/scripts/generate-tool-api-types.mts --spec ../SwaggerDoc.json --output editor/src/core/types/toolApiTypes.ts --check
-
  licenses:generate:
    desc: "Generate frontend license report"
    deps: [install]
    cmds:
      - node editor/scripts/generate-licenses.js
-
-  # ============================================================
-  # Clean
-  # ============================================================
-
-  clean:
-    desc: "Clean build artifacts and caches"
-    cmds:
-      - cmd: powershell rm -Recurse -Force -ErrorAction SilentlyContinue node_modules/.vite, editor/dist, dist
-        platforms: [windows]
-      - cmd: rm -rf node_modules/.vite editor/dist dist
-        platforms: [linux, darwin]
@@ -1,133 +0,0 @@
-version: '3'
-
-# Repo-wide lint/format/secret checks - the single source of truth that the git
-# pre-commit hook (.pre-commit-config.yaml) and CI (pre_commit.yml) both call.
-
-vars:
-  # File selections as git pathspecs: git does the include/exclude matching, so
-  # there is no grep/xargs and it behaves identically on every platform.
-  PY_FILES: >-
-    'scripts/*.py'
-    '.github/scripts/*.py'
-    'app/core/src/main/resources/static/python/*.py'
-    ':(exclude)*split_photos.py'
-  SPELL_FILES: >-
-    '*.html'
-    '*.css'
-    '*.js'
-    '*.py'
-    '*.md'
-    ':(exclude).vscode/*'
-    ':(exclude).devcontainer/*'
-    ':(exclude)app/core/src/main/resources/*'
-    ':(exclude)app/proprietary/src/main/resources/*'
-    ':(exclude)frontend/editor/public/vendor/*'
-    ':(exclude)*Dockerfile*'
-    ':(exclude)*pdfjs*'
-    ':(exclude)*thirdParty*'
-    ':(exclude)*bootstrap*'
-    ':(exclude)*.min.*'
-    ':(exclude)*diff.js'
-  WS_FILES: >-
-    '*.js'
-    '*.java'
-    '*.py'
-    '*.yml'
-    ':(exclude)*pdfjs*'
-    ':(exclude)*thirdParty*'
-    ':(exclude)*bootstrap*'
-    ':(exclude)*.min.*'
-    ':(exclude)*diff.js'
-    ':(exclude).github/workflows/*'
-  LOCALE_TOML: 'frontend/editor/public/locales/*/translation.toml'
-
-  # gitleaks is pinned + checksum-verified by scripts/pre-commit/install_gitleaks.py,
-  # which owns the version and caches the binary here.
-  GITLEAKS_BIN: '.task/bin/gitleaks{{if eq OS "windows"}}.exe{{end}}'
-
-tasks:
-  default:
-    desc: "Check formatting, spelling, and secrets across the repo"
-    cmds:
-      - task: ruff
-      - task: ruff-format
-      - task: codespell
-      - task: gitleaks
-      - task: whitespace
-      - task: toml-sort
-
-  fix:
-    desc: "Auto-fix formatting, spelling, and secrets issues across the repo"
-    cmds:
-      # Auto-fixers first, then the report-only tools (codespell, gitleaks) so a
-      # finding there does not stop the fixers from running.
-      - task: ruff
-        vars: { FIX: '1' }
-      - task: ruff-format
-        vars: { FIX: '1' }
-      - task: whitespace
-        vars: { FIX: '1' }
-      - task: toml-sort
-        vars: { FIX: '1' }
-      - task: codespell
-      - task: gitleaks
-
-  install:
-    desc: "Install the pinned pre-commit Python tools (ruff, codespell, toml-sort)"
-    run: once
-    cmds:
-      - uv sync --project scripts/pre-commit --locked
-    sources:
-      - scripts/pre-commit/uv.lock
-      - scripts/pre-commit/pyproject.toml
-    status:
-      - test -d scripts/pre-commit/.venv
-
-  clean:
-    desc: "Remove the cached gitleaks binary and the tool virtualenv"
-    cmds:
-      - cmd: rm -rf scripts/pre-commit/.venv .task/bin/gitleaks
-        platforms: [linux, darwin]
-      - cmd: cmd /c "rmdir /s /q scripts\pre-commit\.venv & del /q .task\bin\gitleaks.exe"
-        platforms: [windows]
-        ignore_error: true
-
-  # Individual checks (hidden from `task --list`, but callable, e.g.
-  # `task pre-commit:toml-sort FIX=1`). Pass FIX=1 to auto-fix where supported.
-  ruff:
-    deps: [install]
-    cmds:
-      - uv run --project scripts/pre-commit --no-sync ruff check --line-length=127 {{if .FIX}}--fix {{end}}$(git ls-files {{.PY_FILES}})
-
-  ruff-format:
-    deps: [install]
-    cmds:
-      - uv run --project scripts/pre-commit --no-sync ruff format {{if .FIX}}{{else}}--check {{end}}$(git ls-files {{.PY_FILES}})
-
-  codespell:
-    deps: [install]
-    cmds:
-      - uv run --project scripts/pre-commit --no-sync codespell --ignore-words-list=thirdParty,tabEl,tabEls,Sie,ist,fulfilment --quiet-level=2 $(git ls-files {{.SPELL_FILES}})
-
-  toml-sort:
-    deps: [install]
-    cmds:
-      - uv run --project scripts/pre-commit --no-sync toml-sort --all --ignore-case {{if .FIX}}--in-place{{else}}--check{{end}} {{.LOCALE_TOML}}
-
-  whitespace:
-    cmds:
-      - uv run --no-project python scripts/pre-commit/whitespace.py {{if .FIX}}--fix {{end}}{{.WS_FILES}}
-
-  gitleaks:
-    deps: [gitleaks-bin]
-    # Scan staged changes only, matching the old hook: the git-mode fingerprints
-    # in .gitleaksignore (file:rule:line) still apply, and with nothing staged
-    # this is a no-op. Secrets are never auto-fixed, so FIX has no effect.
-    cmds:
-      - "{{.GITLEAKS_BIN}} git --pre-commit --redact --staged --verbose"
-
-  gitleaks-bin:
-    internal: true
-    desc: "Ensure the pinned, checksum-verified gitleaks binary is cached in .task/bin"
-    cmds:
-      - uv run --no-project python scripts/pre-commit/install_gitleaks.py
@@ -200,9 +200,9 @@ const [ToolName] = (props: BaseToolProps) => {
 ```

 ## 5. Add Translations
-Update translation files. **Important: Only update `en-US` files** - other languages are handled separately.
+Update translation files. **Important: Only update `en-GB` files** - other languages are handled separately.

-**File to update:** `frontend/editor/public/locales/en-US/translation.toml`
+**File to update:** `frontend/editor/public/locales/en-GB/translation.toml`

 **Required Translation Keys**:
 ```toml
@@ -251,7 +251,7 @@ Update translation files. **Important: Only update `en-US` files** - other langu
 ```

 **Translation Notes:**
- **Only update `en-US/translation.toml`** - other locale files are managed separately
+- **Only update `en-GB/translation.toml`** - other locale files are managed separately
 - Use descriptive keys that match your component's `t()` calls
 - Include tooltip translations if you created tooltip hooks
 - Add `options.*` keys if your tool has settings with descriptions
@@ -139,8 +139,7 @@ The project structure is defined in `engine/pyproject.toml`. Any new dependencie

 #### Environment Variables
 - All `VITE_*` variables must be declared in the appropriate committed env file:
-  - `frontend/editor/.env` — core and shared vars (base, loaded in every mode)
-  - `frontend/editor/.env.proprietary` — proprietary-only vars, e.g. the admin portal's SaaS/account-link keys (layered on top of `.env` in proprietary mode)
+  - `frontend/editor/.env` — core, proprietary, and shared vars
  - `frontend/editor/.env.saas` — SaaS-only vars (layered on top of `.env` in SaaS mode)
  - `frontend/editor/.env.desktop` — desktop (Tauri)-only vars (layered on top of `.env` in desktop mode)
 - These files are committed to Git and must not contain private keys
@@ -153,7 +152,7 @@ The project structure is defined in `engine/pyproject.toml`. Any new dependencie
 #### Import Paths - CRITICAL
 **ALWAYS use `@app/*` for imports.** Do not use `@core/*` or `@proprietary/*` unless explicitly wrapping/extending a lower layer implementation.

-For a broader explanation of the frontend layering and override architecture, read @frontend/editor/DeveloperGuide.md
+For a broader explanation of the frontend layering and override architecture, see [frontend/editor/DeveloperGuide.md](frontend/editor/DeveloperGuide.md).

 ```typescript
 // ✅ CORRECT - Use @app/* for all imports
@@ -170,31 +169,7 @@ import { useFileContext } from "@proprietary/contexts/FileContext";
 - Building layer-specific override that wraps a lower layer's component
 - Example: `import { AppProviders as CoreAppProviders } from "@core/components/AppProviders"` when creating proprietary/AppProviders.tsx that extends the core version

-The `@app/*` alias automatically resolves to the correct layer based on build target (core/proprietary/saas/desktop/cloud) and handles the fallback cascade — see "Frontend `cloud/` Layer" below for the full per-flavor order.
-
-#### Frontend `cloud/` Layer
-
-`@app/*` resolves through a per-flavor cascade — first existing file wins (shadow/override):
-
- **core** → core
- **proprietary** → proprietary → core
- **saas** → saas → cloud → proprietary → core
- **desktop** → desktop → cloud → proprietary → core
- **cloud** → cloud → proprietary → core
-
-What goes where:
-
- **core** — OSS base.
- **proprietary** — licensed / offline features.
- **cloud** — the SHARED hosted/SaaS experience used by BOTH saas + desktop: PAYG, wallet, plan, billing, usage meters, cloud config/team/onboarding.
- **saas** — web-only: Supabase web auth, AuthCallback, avatar canvas, `window.location`.
- **desktop** — Tauri-only: keyring authService, tauriHttpClient, native files/windows, backend routing.
-
-`cloud/` MUST NOT import `@supabase/*`, `@tauri-apps/*`, raw `fetch`, `window.location`, `localStorage`, `sessionStorage`, or `import.meta.env.VITE_*` (enforced by ESLint). It reaches platform-specific things only via `@app/*` seams: `services/apiClient`, `auth/session.getAccessToken`, `auth/supabase`, `platform/openExternal`, `services/billing`, `hooks/useSaaSMode` — each provided per-platform in `saas/` and `desktop/`.
-
-Rule of thumb — **move, don't copy**: share via `cloud/`, override by shadowing the same `@app/*` path in a leaf (`saas/` or `desktop/`).
-
-**Cloud feature flags on desktop.** The local `AppConfigContext` reads `/api/v1/config/app-config` from the LOCAL bundled backend, so cloud-only flags (`aiEngineEnabled`, `premiumEnabled`, …) are never seen on desktop. To read the cloud's view, use `useSaasAppConfig()` (`desktop/hooks/useSaasAppConfig.ts`, backed by the general `saasAppConfigService` — SaaS-mode-only, public endpoint, native HTTP, 5-min cache). It returns `null` outside SaaS mode, so cloud features stay off in local/self-hosted and the server keeps the on/off switch (no desktop release needed to flip a flag). Gate a feature behind a per-platform seam — e.g. `useAiEngineEnabled()` (core reads `useAppConfig()`, desktop reads `useSaasAppConfig()`) — rather than hardcoding the flag on.
+The `@app/*` alias automatically resolves to the correct layer based on build target (core/proprietary/desktop) and handles the fallback cascade.

 #### Component Override Pattern (Stub/Shadow)
 Use this pattern for desktop-specific or proprietary-specific features WITHOUT runtime checks or conditionals.
@@ -451,9 +426,8 @@ The frontend is organized with a clear separation of concerns:

 ## Translation Rules

- **CRITICAL**: Always update translations in `en-US` only - all other languages (including `en-GB`) are handled separately
+- **CRITICAL**: Always update translations in `en-GB` only, never `en-US`
 - Translation files are located in `frontend/editor/public/locales/`
- After changing any translation file, run `task pre-commit:fix`

 ## Important Notes

@@ -52,17 +52,6 @@ This guide focuses on developing for Stirling 2.0, including both the React fron
 - Rust and Cargo (required for Tauri desktop app development)
 - Tauri CLI (install with `cargo install tauri-cli`)

-### Optional System Dependencies
-
-These are not required to run the app but enable specific features. The app detects them at startup and disables the relevant features if they are missing.
-
-| Dependency | Feature | Install |
-|---|---|---|
-| LibreOffice | File-to-PDF conversions | `brew install libreoffice` / `apt install libreoffice` |
-| Tesseract | OCR | `brew install tesseract` / `apt install tesseract-ocr` |
-| WeasyPrint | AI document creation | `brew install weasyprint` / `apt install weasyprint` |
-| qpdf | PDF optimisation | `brew install qpdf` / `apt install qpdf` |
-
 ### Setup Steps

 1. Clone the repository:
@@ -92,7 +81,7 @@ Visit the [Lombok website](https://projectlombok.org/setup/) for installation in

 5. Add environment variable
 For local testing, you should generally be testing the full 'Security' version of Stirling PDF. To do this, you must add the environment flag DISABLE_ADDITIONAL_FEATURES=false to your system and/or IDE build/run step.
-6. **Frontend Setup (Required for Stirling 2.0)**
+5. **Frontend Setup (Required for Stirling 2.0)**
   Navigate to the frontend directory and install dependencies using npm.

 ### Verify Setup
@@ -275,7 +264,7 @@ Stirling-PDF uses different Docker images for various configurations. The build
 1. Set the security environment variable:

   ```bash
-   export DISABLE_ADDITIONAL_FEATURES=true  # or false to enable login and security features for builds
+   export DISABLE_ADDITIONAL_FEATURES=true  # or false for to enable login and security features for builds
   ```

 2. Build the project:
@@ -305,7 +294,7 @@ Stirling-PDF uses different Docker images for various configurations. The build
   docker build --no-cache --pull --build-arg VERSION_TAG=alpha -t stirlingtools/stirling-pdf:latest-fat -f ./Dockerfile.fat .
   ```

-Note: The `--no-cache` and `--pull` flags ensure that the build process uses the latest base images and doesn't use cached layers, which is useful for testing and ensuring reproducible builds. However, to improve build times these can often be removed depending on your use case
+Note: The `--no-cache` and `--pull` flags ensure that the build process uses the latest base images and doesn't use cached layers, which is useful for testing and ensuring reproducible builds. however to improve build times these can often be removed depending on your usecase

 ## 7. Testing

@@ -587,7 +576,7 @@ When adding a new feature or modifying existing ones in Stirling-PDF, you'll nee
 Find the existing `messages.properties` files in the `stirling-pdf/src/main/resources` directory. You'll see files like:

 - `messages.properties` (default, usually English)
- `messages_en_US.properties`
+- `messages_en_GB.properties`
 - `messages_fr_FR.properties`
 - `messages_de_DE.properties`
 - etc.
@@ -16,12 +16,10 @@ if that directory exists, is licensed under the license defined in "frontend/edi
 if that directory exists, is licensed under the license defined in "frontend/editor/src/desktop/LICENSE".
 * All content that resides under the "frontend/editor/src/saas/" directory of this repository,
 if that directory exists, is licensed under the license defined in "frontend/editor/src/saas/LICENSE".
-* All content that resides under the "frontend/editor/src/cloud/" directory of this repository,
-if that directory exists, is licensed under the license defined in "frontend/editor/src/cloud/LICENSE".
 * All content that resides under the "frontend/editor/src/prototypes/" directory of this repository,
 if that directory exists, is licensed under the license defined in "frontend/editor/src/prototypes/LICENSE".
-* All content that resides under the "frontend/editor/src/portal/" directory of this repository,
-if that directory exists, is licensed under the license defined in "frontend/editor/src/portal/LICENSE".
+* All content that resides under the "frontend/portal/" directory of this repository,
+if that directory exists, is licensed under the license defined in "frontend/portal/LICENSE".
 * Content outside of the above mentioned directories or restrictions above is
 available under the MIT License as defined below.

@@ -53,14 +53,14 @@ For full installation options (including desktop and Kubernetes), see our [Docum

 ## Support

- **Community**: [Discord](https://discord.gg/HYmhKj45pU)
- **Bug Reports**: [GitHub Issues](https://github.com/Stirling-Tools/Stirling-PDF/issues)
+- **Community** [Discord](https://discord.gg/HYmhKj45pU)
+- **Bug Reports**: [Github issues](https://github.com/Stirling-Tools/Stirling-PDF/issues)

 ## Contributing

 We welcome contributions! Please see [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines.

-This project uses [Task](https://taskfile.dev/) as a unified command runner for all build, dev, and test commands. Run `task dev` to get started running the editor, run `task` to see the most common commands, or see the [Developer Guide](DeveloperGuide.md) for full details.
+This project uses [Task](https://taskfile.dev/) as a unified command runner for all build, dev, and test commands. Run `task install` to get started, or see the [Developer Guide](DeveloperGuide.md) for full details.

 For adding translations, see the [Translation Guide](devGuide/HowToAddNewLanguage.md).

@@ -25,28 +25,8 @@ includes:
  e2e:
    taskfile: .taskfiles/e2e.yml
    dir: .
-  pre-commit:
-    taskfile: .taskfiles/pre-commit.yml
-    dir: .

 tasks:
-  # ============================================================
-  # Help (shown when you run `task` with no arguments)
-  # ============================================================
-
-  default:
-    desc: "List the most common commands"
-    silent: true
-    cmds:
-      - |
-        echo "Common commands (run 'task --list' to see all):"
-        echo ""
-        echo "  task dev            Start backend & frontend on free ports"
-        echo "  task backend:dev    Start backend on default port"
-        echo "  task frontend:dev   Start frontend on default port"
-        echo "  task desktop:dev    Start desktop app"
-        echo "  task check          Quality gate (lint, typecheck, test, etc.)"
-
  # ============================================================
  # Setup & Prerequisites
  # ============================================================
@@ -78,41 +58,26 @@ tasks:
          BACKEND_URL: 'http://localhost:{{.BACKEND_PORT}}'
          OPEN: "true"

-  dev:portal:
-    desc: "Start backend + editor; the portal is an admin route at /portal"
+  dev:saas:
+    desc: "Start SaaS backend + frontend concurrently on free ports"
    vars:
      PORTS:
        sh: '{{if eq OS "windows"}}{{.FIND_FREE_PORT_PS}} 8080 5173{{else}}{{.FIND_FREE_PORT_SH}} 8080 5173{{end}}'
      BACKEND_PORT: '{{index (splitList "\n" .PORTS) 0}}'
-      EDITOR_PORT: '{{index (splitList "\n" .PORTS) 1}}'
+      FRONTEND_PORT: '{{index (splitList "\n" .PORTS) 1}}'
    deps:
-      - task: backend:dev
+      - task: backend:dev:saas
        vars:
          PORT: '{{.BACKEND_PORT}}'
-          SECURITY_ENABLELOGIN: "true"
-          POLICIES_ENABLED: "true"
-      - task: frontend:dev:proprietary
+      - task: frontend:dev:saas
        vars:
-          PORT: '{{.EDITOR_PORT}}'
+          PORT: '{{.FRONTEND_PORT}}'
          BACKEND_URL: 'http://localhost:{{.BACKEND_PORT}}'
          OPEN: "true"

-  dev:saas:
-    desc: "Start SaaS backend + frontend concurrently on free ports"
-    cmds:
-      - task: dev:_all
-        vars: { FRONTEND: saas, BACKEND: saas }
-
  dev:all:
    desc: "Start backend + frontend + engine concurrently on free ports"
-    cmds:
-      - task: dev:_all
-
-  dev:_all:
-    internal: true
    vars:
-      FRONTEND: '{{.FRONTEND | default "proprietary"}}'
-      BACKEND: '{{.BACKEND | default "proprietary"}}'
      PORTS:
        sh: '{{if eq OS "windows"}}{{.FIND_FREE_PORT_PS}} 8080 5173 5001{{else}}{{.FIND_FREE_PORT_SH}} 8080 5173 5001{{end}}'
      BACKEND_PORT: '{{index (splitList "\n" .PORTS) 0}}'
@@ -122,12 +87,11 @@ tasks:
      - task: engine:dev
        vars:
          PORT: '{{.ENGINE_PORT}}'
-      - task: 'backend:dev:{{.BACKEND}}'
+      - task: backend:dev
        vars:
          PORT: '{{.BACKEND_PORT}}'
          AIENGINE_URL: 'http://localhost:{{.ENGINE_PORT}}'
-          AIENGINE_ENABLED: "true"
-      - task: 'frontend:dev:{{.FRONTEND}}'
+      - task: frontend:dev
        vars:
          PORT: '{{.FRONTEND_PORT}}'
          BACKEND_URL: 'http://localhost:{{.BACKEND_PORT}}'
@@ -185,16 +149,6 @@ tasks:
      - task: frontend:format:check
      - task: engine:format:check

-  # ============================================================
-  # Code generation
-  # ============================================================
-
-  tool-models:
-    desc: "Generate all API models from the Java OpenAPI spec"
-    cmds:
-      - task: frontend:tool-models
-      - task: engine:tool-models
-
  # ============================================================
  # Quality Gate
  # ============================================================
@@ -221,6 +175,4 @@ tasks:
    desc: "Clean all build artifacts"
    cmds:
      - task: backend:clean
-      - task: frontend:clean
      - task: engine:clean
-      - task: pre-commit:clean
@@ -1,8 +0,0 @@
-# Committed defaults for `task backend:dev:proprietary` (self-hosted / proprietary
-# flavor). Local overrides + secrets live in app/.env.proprietary.local (ignored).
-
-# Combined-billing account link (Mode A). Feature-flagged: OFF until release.
-# Flip to true in app/.env.proprietary.local to test linking locally.
-STIRLING_BILLING_ACCOUNT_LINK_ENABLED=false
-# SaaS base URL the linked instance calls (register + entitlement).
-STIRLING_BILLING_ACCOUNT_LINK_SAAS_BASE_URL=https://stirling.com/app
@@ -1,4 +1,3 @@
 # Whitelist committed env defaults. `.env.saas.local` (and any other .env*)
 # stays ignored via the root .gitignore.
 !.env.saas
-!.env.proprietary
@@ -80,18 +80,10 @@
      "moduleName": ".*",
      "moduleLicense": "Apache License Version 2.0"
    },
-    {
-      "moduleName": ".*",
-      "moduleLicense": "Apache License version 2.0"
-    },
    {
      "moduleName": ".*",
      "moduleLicense": "Apache License, Version 2.0"
    },
-    {
-      "moduleName": ".*",
-      "moduleLicense": "Apache License, version 2.0"
-    },
    {
      "moduleName": ".*",
      "moduleLicense": "The Apache License, Version 2.0"
@@ -116,10 +108,6 @@
      "moduleName": ".*",
      "moduleLicense": "Mozilla Public License 2.0 (MPL-2.0)"
    },
-    {
-      "moduleName": ".*",
-      "moduleLicense": "Mozilla Public License Version 2.0"
-    },
    {
      "moduleName": ".*",
      "moduleLicense": 	"CDDL+GPL License"
@@ -184,14 +172,6 @@
      "moduleName": ".*",
      "moduleLicense": "Eclipse Public License, Version 2.0"
    },
-    {
-      "moduleName": ".*",
-      "moduleLicense": "EPL-2.0"
-    },
-    {
-      "moduleName": ".*",
-      "moduleLicense": "LGPL-2.1-only"
-    },
    {
      "moduleName": ".*",
      "moduleLicense": "Ubuntu Font Licence 1.0"
@@ -29,13 +29,13 @@ spotless {
    }
 }
 dependencies {
-    api "com.google.guava:guava:${guavaVersion}"
+    api 'com.google.guava:guava:33.6.0-jre'
    api 'org.springframework.boot:spring-boot-starter-webmvc'
    api 'org.springframework.boot:spring-boot-starter-aspectj'
    api 'com.googlecode.owasp-java-html-sanitizer:owasp-java-html-sanitizer:20260313.1'
    api 'com.fathzer:javaluator:3.0.6'
    api 'com.posthog.java:posthog:1.2.0'
-    api "org.apache.commons:commons-lang3:${commonsLang3}"
+    api 'org.apache.commons:commons-lang3:3.20.0'
    api 'com.drewnoakes:metadata-extractor:2.20.0' // Image metadata extractor
    api 'com.vladsch.flexmark:flexmark-html2md-converter:0.64.8'
    api "org.apache.pdfbox:pdfbox:$pdfboxVersion"
@@ -60,7 +60,7 @@ dependencies {
        exclude group: 'com.google.code.gson', module: 'gson'
    }

-    api "com.stirling:jpdfium:${jpdfiumVersion}"
+    api 'com.stirling:jpdfium:1.0.2'

    // -PjpdfiumPlatforms=all|<csv of linux-x64,linux-arm64,darwin-x64,darwin-arm64,windows-x64>
    def jpdfiumPlatformsProp = (project.findProperty('jpdfiumPlatforms') ?: 'all').toString().trim()
@@ -75,12 +75,12 @@ dependencies {
    }
    logger.lifecycle("JPDFium native platforms: ${jpdfiumPlatforms.join(', ')}")
    jpdfiumPlatforms.each { platform ->
-        runtimeOnly "com.stirling:jpdfium-natives-${platform}:${jpdfiumVersion}"
+        runtimeOnly "com.stirling:jpdfium-natives-${platform}:1.0.2"
    }

    // Bucket4j (local in-process token bucket for RateLimitStore default impl)
-    implementation "com.bucket4j:bucket4j_jdk17-core:${bucket4jVersion}"
+    implementation 'com.bucket4j:bucket4j_jdk17-core:8.19.0'

    // ArchUnit: enforces module dependency direction (see ArchitectureTest)
-    testImplementation "com.tngtech.archunit:archunit-junit5:${archunitVersion}"
+    testImplementation 'com.tngtech.archunit:archunit-junit5:1.4.2'
 }
@@ -0,0 +1,73 @@
+package stirling.software.SPDF.pdf.parser;
+
+import static stirling.software.SPDF.pdf.parser.PdfModels.*;
+
+import java.io.IOException;
+import java.util.List;
+
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.springframework.context.annotation.Primary;
+import org.springframework.stereotype.Service;
+
+import lombok.RequiredArgsConstructor;
+import lombok.extern.slf4j.Slf4j;
+
+/**
+ * Chains table parsers in priority order: Tabula lattice → Tabula stream → {@link
+ * LineAlignmentTableParser}. The first parser returning a result above {@link
+ * #TABULA_CONFIDENCE_THRESHOLD} wins; results from different parsers are never mixed on one page.
+ */
+@Service
+@Primary
+@RequiredArgsConstructor
+@Slf4j
+public class CompositeTableParser implements TableParser {
+
+    /** Min Tabula confidence to accept results; below this LineAlignment is tried instead. */
+    static final float TABULA_CONFIDENCE_THRESHOLD = 0.5f;
+
+    private final TabulaTableParser tabulaParser;
+    private final LineAlignmentTableParser lineAlignmentParser;
+
+    @Override
+    public List<TableFragment> parse(PDDocument document, RawPage rawPage) throws IOException {
+        // Step 1: Tabula lattice mode (ruled/bordered tables).
+        List<TableFragment> latticeResults = filterConfident(tabulaParser.parse(document, rawPage));
+        if (!latticeResults.isEmpty()) {
+            log.debug(
+                    "Page {}: using Tabula lattice ({} table(s))",
+                    rawPage.pageNumber(),
+                    latticeResults.size());
+            return latticeResults;
+        }
+
+        // Step 2: Tabula stream mode (borderless/whitespace-delimited tables).
+        // parseStream is not on the TableParser interface — this intentionally couples to the
+        // concrete TabulaTableParser since stream mode is a Tabula-specific concept.
+        List<TableFragment> streamResults =
+                filterConfident(tabulaParser.parseStream(document, rawPage));
+        if (!streamResults.isEmpty()) {
+            log.debug(
+                    "Page {}: using Tabula stream ({} table(s))",
+                    rawPage.pageNumber(),
+                    streamResults.size());
+            return streamResults;
+        }
+
+        // Step 3: Geometry-based line-alignment fallback.
+        List<TableFragment> lineResults = lineAlignmentParser.parse(document, rawPage);
+        if (!lineResults.isEmpty()) {
+            log.debug(
+                    "Page {}: using LineAlignment ({} table(s))",
+                    rawPage.pageNumber(),
+                    lineResults.size());
+            return lineResults;
+        }
+
+        return List.of();
+    }
+
+    private List<TableFragment> filterConfident(List<TableFragment> tables) {
+        return tables.stream().filter(t -> t.confidence() >= TABULA_CONFIDENCE_THRESHOLD).toList();
+    }
+}
@@ -0,0 +1,528 @@
+package stirling.software.SPDF.pdf.parser;
+
+import static stirling.software.SPDF.pdf.parser.PdfModels.*;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collections;
+import java.util.Comparator;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+import java.util.Optional;
+import java.util.TreeMap;
+import java.util.regex.Pattern;
+
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.springframework.stereotype.Service;
+
+import lombok.extern.slf4j.Slf4j;
+
+/**
+ * Fallback {@link TableParser} for borderless financial tables using text geometry.
+ *
+ * <p>Identifies "anchor lines" (≥2 numeric tokens), builds a column grid from their right-edge
+ * positions, groups vertically proximate anchor lines into table candidates, then scores each group
+ * on column consistency and anchor density (confidence ceiling 0.85).
+ */
+@Service
+@Slf4j
+public class LineAlignmentTableParser implements TableParser {
+
+    /** Width in points of each column position bucket. */
+    static final float COLUMN_BUCKET_PT = 5f;
+
+    /** Tolerance in buckets when matching a token's right-edge to a confirmed column position. */
+    private static final int COLUMN_MATCH_BUCKETS = 2;
+
+    /** Maximum gap (as a multiple of modal line spacing) before splitting a group. */
+    private static final float MAX_GAP_FACTOR = 2.5f;
+
+    /** Minimum anchor rows (numeric-heavy) to form a valid table. */
+    static final int MIN_TABLE_ROWS = 3;
+
+    /** Minimum confirmed column positions to form a valid table. */
+    static final int MIN_COLUMNS = 2;
+
+    /**
+     * Min fraction of anchor lines a column must appear on to be confirmed (permissive for N/A
+     * rows).
+     */
+    private static final double COLUMN_MIN_FREQUENCY = 0.40;
+
+    /**
+     * Matches financial numeric tokens: integers, decimals, parenthetical negatives, currency,
+     * percent, nil dashes.
+     */
+    private static final Pattern NUMERIC =
+            Pattern.compile("^[\\(\\-\\$£€¥]?\\d[\\d,\\.]*[\\)%]?$|^[-–—]$");
+
+    /**
+     * Lines within this y-distance are merged into one row (restores rows split by LineBuilder's
+     * column-gap logic).
+     */
+    static final float ROW_MERGE_TOLERANCE_PT = 2f;
+
+    // ── public API ───────────────────────────────────────────────────────────────────────────────
+
+    @Override
+    public List<TableFragment> parse(PDDocument document, RawPage rawPage) throws IOException {
+        List<RawLine> lines = rawPage.lines();
+        if (lines.size() < MIN_TABLE_ROWS) return List.of();
+
+        float modalSpacing = computeModalSpacing(lines);
+        List<TokenizedLine> tokenized =
+                mergeCoincidentLines(lines.stream().map(this::tokenize).toList());
+
+        List<TokenizedLine> anchors = tokenized.stream().filter(TokenizedLine::isAnchor).toList();
+
+        if (anchors.size() < MIN_TABLE_ROWS) return List.of();
+
+        List<Float> columnGrid = buildColumnGrid(anchors);
+        if (columnGrid.size() < MIN_COLUMNS) {
+            log.debug(
+                    "Page {}: LineAlignment — fewer than {} confirmed columns, skipping",
+                    rawPage.pageNumber(),
+                    MIN_COLUMNS);
+            return List.of();
+        }
+
+        List<List<TokenizedLine>> groups = groupRows(tokenized, columnGrid, modalSpacing);
+
+        List<TableFragment> results = new ArrayList<>();
+        for (int i = 0; i < groups.size(); i++) {
+            buildFragment(groups.get(i), columnGrid, rawPage.pageNumber(), i)
+                    .ifPresent(results::add);
+        }
+
+        log.debug(
+                "Page {}: LineAlignment detected {} table(s) ({} anchor lines, {} columns)",
+                rawPage.pageNumber(),
+                results.size(),
+                anchors.size(),
+                columnGrid.size());
+        return results;
+    }
+
+    // ── coincident-line merging ──────────────────────────────────────────────────────────────────
+
+    /**
+     * Merges tokenised lines sharing the same y-position into one row, rejoining label/value halves
+     * split by LineBuilder.
+     */
+    List<TokenizedLine> mergeCoincidentLines(List<TokenizedLine> tokenized) {
+        if (tokenized.size() < 2) return tokenized;
+
+        List<TokenizedLine> result = new ArrayList<>();
+        int i = 0;
+
+        while (i < tokenized.size()) {
+            float baseY = tokenized.get(i).line().bounds().y();
+            int j = i + 1;
+            while (j < tokenized.size()
+                    && Math.abs(tokenized.get(j).line().bounds().y() - baseY)
+                            <= ROW_MERGE_TOLERANCE_PT) {
+                j++;
+            }
+
+            if (j == i + 1) {
+                result.add(tokenized.get(i));
+            } else {
+                result.add(mergeGroup(tokenized.subList(i, j)));
+            }
+            i = j;
+        }
+
+        return result;
+    }
+
+    private TokenizedLine mergeGroup(List<TokenizedLine> group) {
+        List<TextFragment> mergedFragments =
+                group.stream()
+                        .flatMap(tl -> tl.line().fragments().stream())
+                        .sorted(Comparator.comparingDouble(f -> f.bounds().x()))
+                        .toList();
+
+        Bounds mergedBounds =
+                group.stream()
+                        .map(tl -> tl.line().bounds())
+                        .reduce(Bounds::merge)
+                        .orElse(group.get(0).line().bounds());
+
+        RawLine mergedLine =
+                new RawLine(
+                        group.get(0).line().lineId(),
+                        mergedFragments,
+                        mergedBounds,
+                        group.get(0).line().pageNumber());
+
+        return tokenize(mergedLine);
+    }
+
+    // ── tokenisation ─────────────────────────────────────────────────────────────────────────────
+
+    /**
+     * Splits fragments into word-level tokens; x-positions are estimated linearly within each
+     * fragment.
+     */
+    TokenizedLine tokenize(RawLine line) {
+        List<LineToken> tokens = new ArrayList<>();
+        for (TextFragment frag : line.fragments()) {
+            tokens.addAll(tokensFromFragment(frag));
+        }
+        List<LineToken> numeric = tokens.stream().filter(LineToken::numeric).toList();
+        return new TokenizedLine(line, tokens, numeric);
+    }
+
+    private List<LineToken> tokensFromFragment(TextFragment frag) {
+        String raw = frag.text();
+        if (raw == null || raw.isBlank()) return List.of();
+
+        float fragX = frag.bounds().x();
+        float fragWidth = frag.bounds().width();
+        int rawLen = raw.length();
+
+        List<LineToken> result = new ArrayList<>();
+        int offset = 0;
+        for (String part : raw.split("\\s+")) {
+            if (part.isEmpty()) {
+                offset++;
+                continue;
+            }
+            int idx = raw.indexOf(part, offset);
+            if (idx < 0) idx = offset;
+
+            float tokenX = rawLen > 0 ? fragX + ((float) idx / rawLen) * fragWidth : fragX;
+            float tokenRight =
+                    rawLen > 0
+                            ? fragX + ((float) (idx + part.length()) / rawLen) * fragWidth
+                            : fragX + fragWidth;
+
+            result.add(new LineToken(part, tokenX, tokenRight, NUMERIC.matcher(part).matches()));
+            offset = idx + part.length();
+        }
+        return result;
+    }
+
+    // ── column grid ──────────────────────────────────────────────────────────────────────────────
+
+    /**
+     * Returns confirmed column right-edge positions — those appearing on ≥ {@value
+     * #COLUMN_MIN_FREQUENCY} × N anchor lines.
+     */
+    private List<Float> buildColumnGrid(List<TokenizedLine> anchors) {
+        // bucket → set of line indices that contributed a numeric token to that bucket
+        Map<Integer, List<Integer>> bucketLines = new HashMap<>();
+        for (int i = 0; i < anchors.size(); i++) {
+            for (LineToken t : anchors.get(i).numeric()) {
+                int bucket = bucket(t.right());
+                bucketLines.computeIfAbsent(bucket, k -> new ArrayList<>()).add(i);
+            }
+        }
+
+        int minHits =
+                Math.max(MIN_TABLE_ROWS, (int) Math.ceil(anchors.size() * COLUMN_MIN_FREQUENCY));
+
+        // Confirmed buckets → average right-edge for that bucket
+        TreeMap<Integer, Float> confirmed = new TreeMap<>();
+        for (Map.Entry<Integer, List<Integer>> entry : bucketLines.entrySet()) {
+            // Count distinct lines
+            long distinctLines = entry.getValue().stream().distinct().count();
+            if (distinctLines >= minHits) {
+                double avg =
+                        entry.getValue().stream()
+                                .distinct() // weight each line equally regardless of token count
+                                .mapToDouble(
+                                        lineIdx ->
+                                                avgRightEdgeForBucket(
+                                                        anchors, lineIdx, entry.getKey()))
+                                .average()
+                                .orElse(entry.getKey() * (double) COLUMN_BUCKET_PT);
+                confirmed.put(entry.getKey(), (float) avg);
+            }
+        }
+
+        return new ArrayList<>(confirmed.values()); // already sorted by bucket (left to right)
+    }
+
+    /**
+     * Returns the average right-edge position of tokens in {@code line} whose bucket matches {@code
+     * targetBucket}, falling back to the bucket's nominal centre when no tokens match.
+     */
+    private double avgRightEdgeForBucket(
+            List<TokenizedLine> anchors, int lineIdx, int targetBucket) {
+        return anchors.get(lineIdx).numeric().stream()
+                .filter(t -> bucket(t.right()) == targetBucket)
+                .mapToDouble(LineToken::right)
+                .average()
+                .orElse(targetBucket * (double) COLUMN_BUCKET_PT);
+    }
+
+    // ── grouping ─────────────────────────────────────────────────────────────────────────────────
+
+    /**
+     * Groups anchor lines into table candidates, including adjacent label rows; a gap &gt;
+     * MAX_GAP_FACTOR × modal spacing splits groups.
+     */
+    private List<List<TokenizedLine>> groupRows(
+            List<TokenizedLine> all, List<Float> columnGrid, float modalSpacing) {
+        float maxGap = modalSpacing > 0 ? modalSpacing * MAX_GAP_FACTOR : 30f;
+
+        List<List<TokenizedLine>> groups = new ArrayList<>();
+        List<TokenizedLine> current = new ArrayList<>();
+
+        for (int i = 0; i < all.size(); i++) {
+            TokenizedLine tl = all.get(i);
+            boolean fits = tl.isAnchor() && matchesGrid(tl, columnGrid);
+
+            if (current.isEmpty()) {
+                if (fits) current.add(tl);
+                continue;
+            }
+
+            float gap =
+                    tl.line().bounds().y()
+                            - current.get(current.size() - 1).line().bounds().bottom();
+
+            if (gap > maxGap) {
+                groups.add(current);
+                current = new ArrayList<>();
+                if (fits) current.add(tl);
+                continue;
+            }
+
+            if (fits) {
+                current.add(tl);
+            } else if (!tl.line().text().isBlank()) {
+                // Include non-anchor lines (labels) only if they have text and are within
+                // proximity.
+                current.add(tl);
+            }
+        }
+
+        if (!current.isEmpty()) groups.add(current);
+
+        return groups.stream().filter(g -> hasEnoughAnchorRows(g, columnGrid)).toList();
+    }
+
+    private boolean hasEnoughAnchorRows(List<TokenizedLine> group, List<Float> columnGrid) {
+        return group.stream().filter(r -> r.isAnchor() && matchesGrid(r, columnGrid)).count()
+                >= MIN_TABLE_ROWS;
+    }
+
+    /** A line "matches" the grid when ≥ 60 % of its numeric tokens land in confirmed columns. */
+    private boolean matchesGrid(TokenizedLine tl, List<Float> columnGrid) {
+        if (tl.numeric().isEmpty()) return false;
+        long matches =
+                tl.numeric().stream()
+                        .filter(t -> nearestColumnIndex(t.right(), columnGrid) >= 0)
+                        .count();
+        return (double) matches / tl.numeric().size() >= 0.60;
+    }
+
+    private boolean hasInconsistentColumnMatch(TokenizedLine tl, List<Float> columnGrid) {
+        if (tl.numeric().isEmpty()) return false;
+        long hits =
+                tl.numeric().stream()
+                        .filter(t -> nearestColumnIndex(t.right(), columnGrid) >= 0)
+                        .count();
+        return (double) hits / tl.numeric().size() < 0.60;
+    }
+
+    // ── fragment assembly ────────────────────────────────────────────────────────────────────────
+
+    private Optional<TableFragment> buildFragment(
+            List<TokenizedLine> group, List<Float> columnGrid, int pageNumber, int tableIndex) {
+
+        long anchorCount =
+                group.stream().filter(r -> r.isAnchor() && matchesGrid(r, columnGrid)).count();
+        if (anchorCount < MIN_TABLE_ROWS) return Optional.empty();
+
+        List<String> warnings = new ArrayList<>();
+        List<List<String>> rawRows = new ArrayList<>();
+        List<TableRow> rows = new ArrayList<>();
+
+        for (int rowIdx = 0; rowIdx < group.size(); rowIdx++) {
+            TokenizedLine tl = group.get(rowIdx);
+            List<String> rawRow = buildRawRow(tl, columnGrid);
+            rawRows.add(Collections.unmodifiableList(rawRow));
+            rows.add(buildTableRow(rowIdx, tl, rawRow, columnGrid));
+        }
+
+        // Column count = 1 label column + confirmed numeric columns
+        int colCount = columnGrid.size() + 1;
+        Bounds bounds = computeGroupBounds(group);
+        float confidence = computeConfidence(group, columnGrid, warnings);
+
+        return Optional.of(
+                new TableFragment(
+                        "tbl-la-p" + pageNumber + "-" + tableIndex,
+                        pageNumber,
+                        bounds,
+                        List.of(),
+                        Collections.unmodifiableList(rows),
+                        Collections.unmodifiableList(rawRows),
+                        colCount,
+                        confidence,
+                        Collections.unmodifiableList(warnings),
+                        null));
+    }
+
+    /**
+     * Builds a raw row as a list of strings: index 0 = label text, indices 1..N = column values.
+     */
+    private List<String> buildRawRow(TokenizedLine tl, List<Float> columnGrid) {
+        String[] cells = new String[columnGrid.size() + 1];
+        Arrays.fill(cells, "");
+
+        // Separate label tokens (those not landing in any confirmed column) from column tokens.
+        List<String> labelParts = new ArrayList<>();
+        for (LineToken token : tl.all()) {
+            int col = nearestColumnIndex(token.right(), columnGrid);
+            if (col >= 0 && token.numeric()) {
+                int cellIdx = col + 1;
+                cells[cellIdx] =
+                        cells[cellIdx].isEmpty()
+                                ? token.text()
+                                : cells[cellIdx] + " " + token.text();
+            } else {
+                labelParts.add(token.text());
+            }
+        }
+        cells[0] = String.join(" ", labelParts).trim();
+        return Arrays.asList(cells);
+    }
+
+    private TableRow buildTableRow(
+            int rowIdx, TokenizedLine tl, List<String> rawRow, List<Float> columnGrid) {
+        List<TableCell> cells = new ArrayList<>(rawRow.size());
+
+        // Label cell: use the line's full bounds as an approximation.
+        cells.add(TableCell.of(0, rawRow.get(0), tl.line().bounds()));
+
+        for (int col = 0; col < columnGrid.size(); col++) {
+            String text = col + 1 < rawRow.size() ? rawRow.get(col + 1) : "";
+            float right = columnGrid.get(col);
+            float left = col > 0 ? columnGrid.get(col - 1) : right - 50f;
+            Bounds cellBounds =
+                    new Bounds(
+                            left,
+                            tl.line().bounds().y(),
+                            right - left,
+                            tl.line().bounds().height());
+            cells.add(TableCell.of(col + 1, text, cellBounds));
+        }
+        return new TableRow(rowIdx, Collections.unmodifiableList(cells));
+    }
+
+    // ── confidence scoring ───────────────────────────────────────────────────────────────────────
+
+    /**
+     * Heuristic score in [0.0, 0.85] (ceiling keeps results below Tabula lattice which starts at
+     * 1.0). Base 0.70; +0.05/col beyond 2 (max +0.10); +0.05 at ≥5 anchors, +0.05 at ≥8; −0.15 if
+     * &gt;30 % of anchors have inconsistent columns; −0.10 if non-anchors outnumber anchors.
+     */
+    private float computeConfidence(
+            List<TokenizedLine> group, List<Float> columnGrid, List<String> warnings) {
+        float score = 0.70f;
+
+        long anchorCount =
+                group.stream().filter(r -> r.isAnchor() && matchesGrid(r, columnGrid)).count();
+        long totalRows = group.size();
+
+        // More columns
+        int extraCols = Math.min(columnGrid.size() - MIN_COLUMNS, 2);
+        score += extraCols * 0.05f;
+
+        // More anchor rows
+        if (anchorCount >= 5) score += 0.05f;
+        if (anchorCount >= 8) score += 0.05f;
+
+        // Inconsistent column matching
+        long inconsistent =
+                group.stream()
+                        .filter(TokenizedLine::isAnchor)
+                        .filter(tl -> hasInconsistentColumnMatch(tl, columnGrid))
+                        .count();
+        if (inconsistent > anchorCount * 0.30) {
+            score -= 0.15f;
+            warnings.add(
+                    "Column match inconsistent on "
+                            + inconsistent
+                            + "/"
+                            + anchorCount
+                            + " anchor rows");
+        }
+
+        // Label-heavy
+        long nonAnchor = totalRows - anchorCount;
+        if (nonAnchor > anchorCount) {
+            score -= 0.10f;
+            warnings.add(
+                    "Non-anchor rows ("
+                            + nonAnchor
+                            + ") outnumber anchor rows ("
+                            + anchorCount
+                            + ")");
+        }
+
+        return Math.max(0f, Math.min(0.85f, score));
+    }
+
+    // ── utility ──────────────────────────────────────────────────────────────────────────────────
+
+    /**
+     * Returns the grid index nearest to {@code rightEdge}, or -1 if none is within {@value
+     * #COLUMN_MATCH_BUCKETS} buckets.
+     */
+    private int nearestColumnIndex(float rightEdge, List<Float> grid) {
+        int nearest = -1;
+        float minDist = COLUMN_MATCH_BUCKETS * COLUMN_BUCKET_PT + 1f;
+        for (int i = 0; i < grid.size(); i++) {
+            float dist = Math.abs(rightEdge - grid.get(i));
+            if (dist < minDist) {
+                minDist = dist;
+                nearest = i;
+            }
+        }
+        return nearest;
+    }
+
+    private Bounds computeGroupBounds(List<TokenizedLine> group) {
+        return group.stream()
+                .map(tl -> tl.line().bounds())
+                .reduce(Bounds::merge)
+                .orElse(new Bounds(0, 0, 0, 0));
+    }
+
+    /** Modal gap between consecutive line edges, used to calibrate the group-split threshold. */
+    private float computeModalSpacing(List<RawLine> lines) {
+        if (lines.size() < 2) return 0f;
+        Map<Float, Long> freq = new HashMap<>();
+        for (int i = 1; i < lines.size(); i++) {
+            float gap = lines.get(i).bounds().y() - lines.get(i - 1).bounds().bottom();
+            if (gap > 0) freq.merge(Math.round(gap / 2f) * 2f, 1L, Long::sum);
+        }
+        return freq.entrySet().stream()
+                .max(Map.Entry.comparingByValue())
+                .map(Map.Entry::getKey)
+                .orElse(0f);
+    }
+
+    private static int bucket(float x) {
+        return Math.round(x / COLUMN_BUCKET_PT);
+    }
+
+    // ── private data types ───────────────────────────────────────────────────────────────────────
+
+    /** A word-level token with an approximate right-edge x-position. */
+    record LineToken(String text, float x, float right, boolean numeric) {}
+
+    /** A {@link RawLine} with tokens pre-computed; an "anchor" has ≥ 2 numeric tokens. */
+    record TokenizedLine(RawLine line, List<LineToken> all, List<LineToken> numeric) {
+        boolean isAnchor() {
+            return numeric.size() >= 2;
+        }
+    }
+}
@@ -0,0 +1,139 @@
+package stirling.software.SPDF.pdf.parser;
+
+import static stirling.software.SPDF.pdf.parser.PdfModels.*;
+
+import java.util.ArrayList;
+import java.util.Comparator;
+import java.util.List;
+
+import org.springframework.stereotype.Service;
+
+import lombok.extern.slf4j.Slf4j;
+
+/**
+ * Groups {@link TextFragment} objects into visual {@link RawLine}s using baseline proximity.
+ *
+ * <p>Fragments are on the same line when their baselines are within a font-size-derived tolerance.
+ * A new line starts whenever the horizontal gap exceeds an adaptive column-gap threshold ({@code
+ * max(effectiveWidth * COLUMN_GAP_RATIO, COLUMN_GAP_MIN_PT)}), splitting two-column text.
+ */
+@Service
+@Slf4j
+public class LineBuilder {
+
+    /** Baseline tolerance as a fraction of font size; 0.5 keeps mixed-size text on one line. */
+    private static final float BASELINE_TOLERANCE_FACTOR = 0.5f;
+
+    /** Absolute minimum tolerance so tiny font sizes don't collapse multi-line content. */
+    private static final float MIN_BASELINE_TOLERANCE = 2f;
+
+    /**
+     * Column-gap threshold as a fraction of page width; 0.10 clears tab stops but stays below
+     * two-column gutters.
+     */
+    static final float COLUMN_GAP_RATIO = 0.10f;
+
+    /** Floor for the column-gap threshold so narrow pages don't over-split lines. */
+    static final float COLUMN_GAP_MIN_PT = 40f;
+
+    public List<RawLine> build(List<TextFragment> fragments, int pageNumber) {
+        if (fragments.isEmpty()) return List.of();
+
+        float effectiveWidth = inferEffectiveWidth(fragments);
+        float columnGapThreshold = Math.max(effectiveWidth * COLUMN_GAP_RATIO, COLUMN_GAP_MIN_PT);
+        log.debug(
+                "LineBuilder page {}: effectiveWidth={:.1f}pt, columnGapThreshold={:.1f}pt",
+                pageNumber,
+                effectiveWidth,
+                columnGapThreshold);
+
+        // Sort top-to-bottom first, then left-to-right within the same baseline band.
+        List<TextFragment> sorted =
+                fragments.stream()
+                        .sorted(
+                                Comparator.comparingDouble(TextFragment::baseline)
+                                        .thenComparingDouble(f -> f.bounds().x()))
+                        .toList();
+
+        List<List<TextFragment>> groups = groupByBaseline(sorted, columnGapThreshold);
+
+        List<RawLine> lines = new ArrayList<>(groups.size());
+        for (int i = 0; i < groups.size(); i++) {
+            List<TextFragment> group =
+                    groups.get(i).stream()
+                            .sorted(Comparator.comparingDouble(f -> f.bounds().x()))
+                            .toList();
+
+            Bounds lineBounds =
+                    group.stream()
+                            .map(TextFragment::bounds)
+                            .reduce(Bounds::merge)
+                            .orElse(new Bounds(0, 0, 0, 0));
+
+            lines.add(new RawLine("ln-p" + pageNumber + "-" + i, group, lineBounds, pageNumber));
+        }
+        return lines;
+    }
+
+    private List<List<TextFragment>> groupByBaseline(
+            List<TextFragment> sorted, float columnGapThreshold) {
+        List<List<TextFragment>> groups = new ArrayList<>();
+        List<TextFragment> current = new ArrayList<>();
+        float currentBaseline = Float.NaN;
+
+        for (TextFragment fragment : sorted) {
+            if (current.isEmpty()) {
+                current.add(fragment);
+                currentBaseline = fragment.baseline();
+                continue;
+            }
+
+            float maxFontSize =
+                    Math.max(
+                            fragment.fontSize(),
+                            (float)
+                                    current.stream()
+                                            .mapToDouble(TextFragment::fontSize)
+                                            .max()
+                                            .orElse(0));
+            float tolerance =
+                    Math.max(maxFontSize * BASELINE_TOLERANCE_FACTOR, MIN_BASELINE_TOLERANCE);
+
+            boolean sameBaseline = Math.abs(fragment.baseline() - currentBaseline) <= tolerance;
+            boolean columnGap = sameBaseline && hasColumnGap(fragment, current, columnGapThreshold);
+
+            if (sameBaseline && !columnGap) {
+                current.add(fragment);
+                // Anchor to the weighted mean baseline so long lines stay stable.
+                currentBaseline =
+                        (currentBaseline * (current.size() - 1) + fragment.baseline())
+                                / current.size();
+            } else {
+                groups.add(current);
+                current = new ArrayList<>();
+                current.add(fragment);
+                currentBaseline = fragment.baseline();
+            }
+        }
+
+        if (!current.isEmpty()) groups.add(current);
+        return groups;
+    }
+
+    /**
+     * True when the gap from the rightmost fragment in {@code group} to {@code next} exceeds {@code
+     * threshold}.
+     */
+    private static boolean hasColumnGap(
+            TextFragment next, List<TextFragment> group, float threshold) {
+        float lastRight = group.get(group.size() - 1).bounds().right();
+        return next.bounds().x() - lastRight > threshold;
+    }
+
+    /** Infers effective page width from the rightmost fragment right-edge plus a 10 % margin. */
+    private static float inferEffectiveWidth(List<TextFragment> fragments) {
+        double maxRight =
+                fragments.stream().mapToDouble(f -> f.bounds().right()).max().orElse(500.0);
+        return (float) maxRight * 1.10f;
+    }
+}
@@ -0,0 +1,79 @@
+package stirling.software.SPDF.pdf.parser;
+
+import static stirling.software.SPDF.pdf.parser.PdfModels.*;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.List;
+
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.common.PDRectangle;
+import org.springframework.stereotype.Service;
+
+import lombok.RequiredArgsConstructor;
+import lombok.extern.slf4j.Slf4j;
+
+/**
+ * Runs the per-page ingestion pipeline: {@link WordExtractingStripper} → {@link LineBuilder} →
+ * {@link TableParser}, producing a {@link PdfModels.ParsedPage} per page. The caller owns the
+ * {@link PDDocument} lifecycle.
+ */
+@Service
+@RequiredArgsConstructor
+@Slf4j
+public class PdfIngester {
+
+    private final LineBuilder lineBuilder;
+    private final TableParser tableParser;
+
+    public List<ParsedPage> parse(PDDocument document) throws IOException {
+        return parse(document, document.getNumberOfPages());
+    }
+
+    public List<ParsedPage> parse(PDDocument document, int maxPages) throws IOException {
+        int pageCount = Math.min(document.getNumberOfPages(), maxPages);
+        List<ParsedPage> pages = new ArrayList<>(pageCount);
+        long fragmentsMs = 0;
+        long tablesMs = 0;
+        long t0 = System.currentTimeMillis();
+
+        for (int p = 1; p <= pageCount; p++) {
+            long ft = System.currentTimeMillis();
+            List<TextFragment> fragments = extractFragments(document, p);
+            fragmentsMs += System.currentTimeMillis() - ft;
+
+            PDPage page = document.getPage(p - 1);
+            PDRectangle mediaBox = page.getMediaBox();
+            List<RawLine> lines = lineBuilder.build(fragments, p);
+            RawPage rawPage = new RawPage(p, mediaBox.getWidth(), mediaBox.getHeight(), lines);
+
+            long tt = System.currentTimeMillis();
+            List<TableFragment> tables = tableParser.parse(document, rawPage);
+            tablesMs += System.currentTimeMillis() - tt;
+
+            log.debug(
+                    "Page {}: {} fragments → {} lines, {} table(s)",
+                    p,
+                    fragments.size(),
+                    lines.size(),
+                    tables.size());
+            pages.add(new ParsedPage(p, mediaBox.getWidth(), mediaBox.getHeight(), tables, lines));
+        }
+
+        log.info(
+                "[timing] parse pages={} total={}ms fragments={}ms tables={}ms",
+                pageCount,
+                System.currentTimeMillis() - t0,
+                fragmentsMs,
+                tablesMs);
+        return pages;
+    }
+
+    private List<TextFragment> extractFragments(PDDocument document, int pageNumber)
+            throws IOException {
+        WordExtractingStripper stripper = new WordExtractingStripper(pageNumber);
+        stripper.getText(document);
+        return stripper.getFragments();
+    }
+}
@@ -0,0 +1,113 @@
+package stirling.software.SPDF.pdf.parser;
+
+import static stirling.software.SPDF.pdf.parser.PdfModels.*;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Collections;
+import java.util.List;
+
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.font.PDFont;
+import org.apache.pdfbox.text.PDFTextStripper;
+import org.apache.pdfbox.text.TextPosition;
+
+/**
+ * Extends {@link PDFTextStripper} to capture per-fragment geometry and font metadata.
+ *
+ * <p>Overrides {@link #writeString} to split each content-stream string into word-level {@link
+ * TextFragment}s with bounding boxes, baseline, font name, and bold flag. Coordinates are in
+ * PDFTextStripper space: (0,0) top-left, Y increases downward, {@code getY()} is the baseline.
+ */
+class WordExtractingStripper extends PDFTextStripper {
+
+    private final int targetPage;
+    private final List<TextFragment> fragments = new ArrayList<>();
+    private int fragmentIndex = 0;
+
+    WordExtractingStripper(int pageNumber) throws IOException {
+        this.targetPage = pageNumber;
+        setStartPage(pageNumber);
+        setEndPage(pageNumber);
+        setSortByPosition(true);
+    }
+
+    @Override
+    protected void startPage(PDPage page) throws IOException {
+        super.startPage(page);
+        fragments.clear();
+        fragmentIndex = 0;
+    }
+
+    @Override
+    protected void writeString(String text, List<TextPosition> textPositions) throws IOException {
+        if (text == null || text.isBlank()) return;
+
+        // Fast path: no whitespace → emit one fragment (most financial PDFs have each
+        // number as its own string operation, so this is the common case).
+        if (text.indexOf(' ') < 0) {
+            emitFragment(text, textPositions);
+            return;
+        }
+
+        // Per-word splitting requires 1:1 text-char to TextPosition correspondence.
+        // Fall back to one fragment when sizes differ (ligatures, encoding edge cases).
+        if (textPositions.size() != text.length()) {
+            emitFragment(text, textPositions);
+            return;
+        }
+
+        // Emit one TextFragment per whitespace-delimited word with accurate per-word bounds.
+        int start = 0;
+        for (int i = 0; i <= text.length(); i++) {
+            if (i == text.length() || text.charAt(i) == ' ') {
+                if (start < i) {
+                    emitFragment(text.substring(start, i), textPositions.subList(start, i));
+                }
+                start = i + 1;
+            }
+        }
+    }
+
+    private void emitFragment(String text, List<TextPosition> positions) {
+        if (positions.isEmpty()) return;
+
+        float minX = Float.MAX_VALUE;
+        float minY = Float.MAX_VALUE;
+        float maxRight = -Float.MAX_VALUE;
+        float maxBaseline = -Float.MAX_VALUE;
+        TextPosition first = null;
+
+        for (TextPosition tp : positions) {
+            if (tp == null) continue;
+            if (first == null) first = tp;
+
+            float x = tp.getX();
+            // getY() is the baseline; top of character = getY() - getHeight().
+            float top = tp.getY() - tp.getHeight();
+            float right = x + tp.getWidth();
+            float baseline = tp.getY();
+
+            minX = Math.min(minX, x);
+            minY = Math.min(minY, top);
+            maxRight = Math.max(maxRight, right);
+            maxBaseline = Math.max(maxBaseline, baseline);
+        }
+
+        if (first == null) return;
+
+        PDFont font = first.getFont();
+        String fontName = font != null ? font.getName() : "";
+        boolean bold = fontName != null && fontName.toLowerCase().contains("bold");
+        // getHeight() gives the rendered glyph height, which is the most reliable visual size.
+        float fontSize = first.getHeight();
+
+        Bounds bounds = new Bounds(minX, minY, maxRight - minX, maxBaseline - minY);
+        String id = "tf-p" + targetPage + "-" + fragmentIndex++;
+        fragments.add(new TextFragment(id, text, bounds, maxBaseline, fontSize, fontName, bold));
+    }
+
+    List<TextFragment> getFragments() {
+        return Collections.unmodifiableList(fragments);
+    }
+}
@@ -11,37 +11,21 @@ public interface FileStore {
    /** Stored file record. */
    record Stored(String fileId, long size) {}

-    /**
-     * Store the given stream and return a generated file id and total bytes written. {@code owner}
-     * may be null to indicate the file has no associated user (anonymous / desktop / async job with
-     * no propagated security context); a non-null value is persisted alongside the data so {@link
-     * #getOwner(String)} can return it later for authorization checks.
-     */
-    Stored store(InputStream in, String originalName, String owner) throws IOException;
-
-    /** Store with no owner. Equivalent to {@link #store(InputStream, String, String)} with null. */
-    default Stored store(InputStream in, String originalName) throws IOException {
-        return store(in, originalName, null);
-    }
+    /** Store the given stream and return a generated file id and total bytes written. */
+    Stored store(InputStream in, String originalName) throws IOException;

    /**
     * Store the file at {@code source} and return a generated file id and total bytes written.
     *
     * <p>Default implementation opens {@code source} as a stream and delegates to {@link
-     * #store(InputStream, String, String)}. Local-disk implementations should override to use a
-     * direct file-to-file copy ({@code Files.copy(source, dest)} can use {@code sendfile(2)} on
-     * Linux), which avoids the two-memory-copy hit of streaming a disk-backed upload through the
-     * JVM heap.
+     * #store(InputStream, String)}. Local-disk implementations should override to use a direct
+     * file-to-file copy ({@code Files.copy(source, dest)} can use {@code sendfile(2)} on Linux),
+     * which avoids the two-memory-copy hit of streaming a disk-backed upload through the JVM heap.
     */
-    default Stored store(Path source, String originalName, String owner) throws IOException {
-        try (InputStream in = Files.newInputStream(source)) {
-            return store(in, originalName, owner);
-        }
-    }
-
-    /** Store with no owner. Equivalent to {@link #store(Path, String, String)} with null. */
    default Stored store(Path source, String originalName) throws IOException {
-        return store(source, originalName, null);
+        try (InputStream in = Files.newInputStream(source)) {
+            return store(in, originalName);
+        }
    }

    /** Open the stored file for streaming reads. Caller closes. */
@@ -58,12 +42,4 @@ public interface FileStore {

    /** Whether the file id exists in the store. */
    boolean exists(String fileId);
-
-    /**
-     * Returns the owner identifier recorded at store time, or {@code null} if the file does not
-     * exist or was stored without an owner. Implementations must not throw when the file is missing
-     * or when the owner record is absent; they should return null so callers can treat "no owner"
-     * as a non-authoritative case.
-     */
-    String getOwner(String fileId) throws IOException;
 }
@@ -3,12 +3,9 @@ package stirling.software.common.cluster.inprocess;
 import java.io.BufferedInputStream;
 import java.io.IOException;
 import java.io.InputStream;
-import java.nio.charset.StandardCharsets;
 import java.nio.file.Files;
 import java.nio.file.Path;
 import java.util.UUID;
-import java.util.concurrent.locks.ReentrantLock;
-import java.util.regex.Pattern;

 import lombok.extern.slf4j.Slf4j;

@@ -18,47 +15,33 @@ import stirling.software.common.cluster.FileStore;
@Slf4j
 public class LocalDiskFileStore implements FileStore {

-    private static final String OWNER_SUFFIX = ".owner";
-
-    // File ids are generated as random UUIDs; reject anything else so a tainted id can never reach
-    // Files.* APIs (defence in depth on top of the resolve() prefix check, and silences CodeQL's
-    // path-injection finding on the resolveOwner sidecar lookup).
-    private static final Pattern UUID_PATTERN =
-            Pattern.compile(
-                    "^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$");
-
    private final String baseDirPath;
-    // Fixed-size lock stripes so concurrent store/delete on the same (or colliding) fileId
-    // serialise the data-file + owner-sidecar pair as one critical section. Striped (not
-    // per-id) so the map never has to be cleaned up; collisions across unrelated ids are
-    // harmless contention.
-    private static final int LOCK_STRIPES = 64;
-    private final ReentrantLock[] stripes = new ReentrantLock[LOCK_STRIPES];

    public LocalDiskFileStore(String baseDirPath) {
        this.baseDirPath = baseDirPath;
-        for (int i = 0; i < LOCK_STRIPES; i++) {
-            stripes[i] = new ReentrantLock();
-        }
    }

    @Override
-    public Stored store(InputStream in, String originalName, String owner) throws IOException {
+    public Stored store(InputStream in, String originalName) throws IOException {
        String fileId = UUID.randomUUID().toString();
        Path filePath = resolve(fileId);
        Files.createDirectories(filePath.getParent());
-        ReentrantLock lock = acquire(fileId);
        boolean success = false;
        try {
            long size = Files.copy(in, filePath);
-            writeOwner(fileId, owner);
            success = true;
            return new Stored(fileId, size);
        } finally {
            if (!success) {
-                cleanupAfterFailedStore(fileId, filePath);
+                try {
+                    Files.deleteIfExists(filePath);
+                } catch (IOException cleanupEx) {
+                    log.warn(
+                            "Failed to clean up partial file {} after store failure",
+                            filePath,
+                            cleanupEx);
+                }
            }
-            release(fileId, lock);
        }
    }

@@ -69,44 +52,27 @@ public class LocalDiskFileStore implements FileStore {
     * the source size before copying so the post-copy stat is unnecessary.
     */
    @Override
-    public Stored store(Path source, String originalName, String owner) throws IOException {
+    public Stored store(Path source, String originalName) throws IOException {
        String fileId = UUID.randomUUID().toString();
        Path filePath = resolve(fileId);
        Files.createDirectories(filePath.getParent());
        long size = Files.size(source);
-        ReentrantLock lock = acquire(fileId);
        boolean success = false;
        try {
            Files.copy(source, filePath);
-            writeOwner(fileId, owner);
            success = true;
            return new Stored(fileId, size);
        } finally {
            if (!success) {
-                cleanupAfterFailedStore(fileId, filePath);
+                try {
+                    Files.deleteIfExists(filePath);
+                } catch (IOException cleanupEx) {
+                    log.warn(
+                            "Failed to clean up partial file {} after store failure",
+                            filePath,
+                            cleanupEx);
+                }
            }
-            release(fileId, lock);
-        }
-    }
-
-    private void writeOwner(String fileId, String owner) throws IOException {
-        if (owner == null || owner.isBlank()) {
-            return;
-        }
-        Path ownerPath = resolveOwner(fileId);
-        Files.write(ownerPath, owner.getBytes(StandardCharsets.UTF_8));
-    }
-
-    private void cleanupAfterFailedStore(String fileId, Path filePath) {
-        try {
-            Files.deleteIfExists(filePath);
-        } catch (IOException cleanupEx) {
-            log.warn("Failed to clean up partial file {} after store failure", filePath, cleanupEx);
-        }
-        try {
-            Files.deleteIfExists(resolveOwner(fileId));
-        } catch (IOException cleanupEx) {
-            log.warn("Failed to clean up owner sidecar for {} after store failure", fileId);
        }
    }

@@ -135,26 +101,11 @@ public class LocalDiskFileStore implements FileStore {

    @Override
    public boolean delete(String fileId) {
-        ReentrantLock lock = acquire(fileId);
        try {
-            // Data first, owner second: a concurrent retrieve that observes the transient
-            // (data-gone, owner-still-present) window simply fails with IOException; the inverse
-            // order would briefly look like an unowned file and could grant cross-user access.
-            boolean removed;
-            try {
-                removed = Files.deleteIfExists(resolve(fileId));
-            } catch (IOException e) {
-                log.error("Error deleting file with ID: {}", fileId, e);
-                return false;
-            }
-            try {
-                Files.deleteIfExists(resolveOwner(fileId));
-            } catch (IOException e) {
-                log.warn("Error deleting owner sidecar for file ID: {}", fileId, e);
-            }
-            return removed;
-        } finally {
-            release(fileId, lock);
+            return Files.deleteIfExists(resolve(fileId));
+        } catch (IOException e) {
+            log.error("Error deleting file with ID: {}", fileId, e);
+            return false;
        }
    }

@@ -163,21 +114,8 @@ public class LocalDiskFileStore implements FileStore {
        return Files.exists(resolve(fileId));
    }

-    @Override
-    public String getOwner(String fileId) throws IOException {
-        Path ownerPath = resolveOwner(fileId);
-        if (!Files.exists(ownerPath)) {
-            return null;
-        }
-        byte[] bytes = Files.readAllBytes(ownerPath);
-        if (bytes.length == 0) {
-            return null;
-        }
-        return new String(bytes, StandardCharsets.UTF_8);
-    }
-
    public Path resolve(String fileId) {
-        if (fileId == null || !UUID_PATTERN.matcher(fileId).matches()) {
+        if (fileId.contains("..") || fileId.contains("/") || fileId.contains("\\")) {
            throw new IllegalArgumentException("Invalid file ID");
        }
        Path basePath = Path.of(baseDirPath).normalize().toAbsolutePath();
@@ -187,19 +125,4 @@ public class LocalDiskFileStore implements FileStore {
        }
        return resolvedPath;
    }
-
-    private Path resolveOwner(String fileId) {
-        Path data = resolve(fileId);
-        return data.resolveSibling(data.getFileName().toString() + OWNER_SUFFIX);
-    }
-
-    private ReentrantLock acquire(String fileId) {
-        ReentrantLock lock = stripes[(fileId.hashCode() & Integer.MAX_VALUE) % LOCK_STRIPES];
-        lock.lock();
-        return lock;
-    }
-
-    private void release(String fileId, ReentrantLock lock) {
-        lock.unlock();
-    }
 }
@@ -3,6 +3,7 @@ package stirling.software.common.configuration;
 import java.io.IOException;
 import java.nio.file.Files;
 import java.nio.file.Path;
+import java.nio.file.Paths;
 import java.util.List;
 import java.util.Locale;
 import java.util.Properties;
@@ -121,18 +122,18 @@ public class AppConfig {

    @Bean(name = "RunningInDocker")
    public boolean runningInDocker() {
-        return Files.exists(Path.of("/.dockerenv"));
+        return Files.exists(Paths.get("/.dockerenv"));
    }

    @Bean(name = "configDirMounted")
    public boolean isRunningInDockerWithConfig() {
-        Path dockerEnv = Path.of("/.dockerenv");
+        Path dockerEnv = Paths.get("/.dockerenv");
        // default to true if not docker
        if (!Files.exists(dockerEnv)) {
            return true;
        }
-        Path mountInfo = Path.of("/proc/1/mountinfo");
-        // this should always exist, if not some unknown use case
+        Path mountInfo = Paths.get("/proc/1/mountinfo");
+        // this should always exist, if not some unknown usecase
        if (!Files.exists(mountInfo)) {
            return true;
        }
@@ -7,6 +7,7 @@ import java.net.URISyntaxException;
 import java.net.URL;
 import java.nio.file.Files;
 import java.nio.file.Path;
+import java.nio.file.Paths;
 import java.nio.file.StandardCopyOption;
 import java.util.List;

@@ -26,7 +27,7 @@ public class ConfigInitializer {

    public void ensureConfigExists() throws IOException, URISyntaxException {
        // 1) If settings file doesn't exist, create from template
-        Path destPath = Path.of(InstallationPathConfig.getSettingsPath());
+        Path destPath = Paths.get(InstallationPathConfig.getSettingsPath());

        boolean settingsFileExists = Files.exists(destPath);

@@ -38,7 +39,7 @@ public class ConfigInitializer {
            if (settingsFileExists) {
                // move settings.yml to settings.yml.{timestamp}.bak
                Path backupPath =
-                        Path.of(
+                        Paths.get(
                                InstallationPathConfig.getSettingsPath()
                                        + "."
                                        + System.currentTimeMillis()
@@ -95,7 +96,7 @@ public class ConfigInitializer {
        }

        // 3) Ensure custom settings file exists
-        Path customSettingsPath = Path.of(InstallationPathConfig.getCustomSettingsPath());
+        Path customSettingsPath = Paths.get(InstallationPathConfig.getCustomSettingsPath());
        if (Files.notExists(customSettingsPath)) {
            Files.createFile(customSettingsPath);
            log.info("Created custom_settings file: {}", customSettingsPath);
@@ -3,6 +3,7 @@ package stirling.software.common.configuration;
 import java.nio.file.Files;
 import java.nio.file.InvalidPathException;
 import java.nio.file.Path;
+import java.nio.file.Paths;
 import java.util.ArrayList;
 import java.util.Collections;
 import java.util.LinkedHashSet;
@@ -200,7 +201,7 @@ public class RuntimePathConfig {

            try {
                // Normalize to absolute path
-                Path path = Path.of(pathStr.trim()).toAbsolutePath().normalize();
+                Path path = Paths.get(pathStr.trim()).toAbsolutePath().normalize();
                String normalizedPath = path.toString();

                // Check for duplicates
@@ -223,9 +224,9 @@ public class RuntimePathConfig {

    private void detectOverlappingPaths(List<String> paths) {
        for (int i = 0; i < paths.size(); i++) {
-            Path path1 = Path.of(paths.get(i));
+            Path path1 = Paths.get(paths.get(i));
            for (int j = i + 1; j < paths.size(); j++) {
-                Path path2 = Path.of(paths.get(j));
+                Path path2 = Paths.get(paths.get(j));

                // Check if one path is a parent of the other
                if (path1.startsWith(path2)) {
@@ -245,10 +246,10 @@ public class RuntimePathConfig {

    private void validatePipelinePaths() {
        try {
-            Path finishedPath = Path.of(pipelineFinishedFoldersPath).toAbsolutePath().normalize();
+            Path finishedPath = Paths.get(pipelineFinishedFoldersPath).toAbsolutePath().normalize();

            for (String watchedPathStr : pipelineWatchedFoldersPaths) {
-                Path watchedPath = Path.of(watchedPathStr).toAbsolutePath().normalize();
+                Path watchedPath = Paths.get(watchedPathStr).toAbsolutePath().normalize();

                // Check if watched folder is same as finished folder
                if (watchedPath.equals(finishedPath)) {
@@ -77,10 +77,8 @@ public class ApplicationProperties {
    private ProcessExecutor processExecutor = new ProcessExecutor();
    private PdfEditor pdfEditor = new PdfEditor();
    private AiEngine aiEngine = new AiEngine();
-    private Mcp mcp = new Mcp();
    private InternalApi internalApi = new InternalApi();
    private Cluster cluster = new Cluster();
-    private Policies policies = new Policies();

    @Bean
    public PropertySource<?> dynamicYamlPropertySource(ConfigurableEnvironment environment)
@@ -204,50 +202,6 @@ public class ApplicationProperties {
        }
    }

-    @Data
-    public static class Policies {
-        /**
-         * Master switch for the policy + sources subsystem (the PAYG-metered automation surface).
-         */
-        private boolean enabled = false;
-
-        /**
-         * Absolute directories that policy folder input sources and output sinks may read from or
-         * write to. Empty (the default) disables folder access entirely, so a policy can never be
-         * pointed at an arbitrary server path. Stirling's own config directory is always
-         * off-limits, and folder access is always disabled in SaaS mode regardless of this list.
-         */
-        private List<String> allowedFolderRoots = new java.util.ArrayList<>();
-
-        /** How often (seconds) the schedule trigger checks for policies whose schedule is due. */
-        private long scheduleSweepSeconds = 60;
-
-        /**
-         * How often (seconds) the folder-watch trigger reconciles its watch registrations and
-         * re-runs every folder-watch policy as a safety net for filesystem events that were missed
-         * (NFS, bind mounts, inotify-queue overflow).
-         */
-        private long watchReconcileSeconds = 300;
-
-        /**
-         * How long (milliseconds) the folder-watch trigger keeps draining filesystem events after
-         * the first, so a burst from a single file copy coalesces into one run instead of many.
-         */
-        private long watchQuietPeriodMs = 500;
-
-        /**
-         * SSE emitter timeout (milliseconds) for streamed runs; generous for long multi-step runs.
-         */
-        private long streamTimeoutMs = 1800000;
-
-        /**
-         * How long (minutes) a finished run's in-memory state is retained before eviction,
-         * mirroring the job-result expiry so rich run state does not outlive the process. Active
-         * and paused runs are kept regardless of age.
-         */
-        private int runExpiryMinutes = 30;
-    }
-
    @Data
    public static class PdfEditor {
        private Cache cache = new Cache();
@@ -302,103 +256,6 @@ public class ApplicationProperties {
        private int longRunningTimeoutSeconds = 600;
    }

-    /**
-     * Model Context Protocol (MCP) server configuration. All keys live under the top-level {@code
-     * mcp.*} prefix. {@link #enabled} defaults to {@code false}: when off, no MCP beans are wired,
-     * no /mcp endpoint exists, and no protected-resource metadata is published.
-     */
-    @Data
-    public static class Mcp {
-
-        /** Master switch. When {@code false} (default), no MCP beans are wired. */
-        private boolean enabled = false;
-
-        /**
-         * When {@code true} (default), invocations require an OAuth scope: {@code mcp.tools.read}
-         * for read-style operations and {@code mcp.tools.write} for write/destructive ones. When
-         * {@code false}, scope checks are skipped (use only if your IdP issues a single coarse
-         * scope).
-         */
-        private boolean scopesEnabled = true;
-
-        /** How often to refresh the AI capabilities manifest from the engine. */
-        private int engineCapabilityRefreshMinutes = 5;
-
-        /**
-         * Tool allow-list (operation ids, e.g. {@code compress-pdf}). When non-empty, ONLY these
-         * operations are exposed over MCP; everything else is hidden, undescribable, and
-         * uninvocable - on top of the global endpoint enable/disable config. Empty = allow all.
-         */
-        private List<String> allowedOperations = new ArrayList<>();
-
-        /**
-         * Tool deny-list (operation ids). Any operation listed here is removed from MCP even if it
-         * would otherwise be allowed. Applied after {@link #allowedOperations}.
-         */
-        private List<String> blockedOperations = new ArrayList<>();
-
-        /** Max MCP request body size in bytes; inline file uploads ride in the JSON-RPC body. */
-        private long maxRequestBytes = 10L * 1024 * 1024;
-
-        /** Results up to this size return inline as base64; larger ones return a fileId only. */
-        private long maxInlineResponseBytes = 10L * 1024 * 1024;
-
-        private Auth auth = new Auth();
-
-        @Data
-        public static class Auth {
-            /**
-             * Authentication mode for the MCP endpoint. {@code oauth} (default) runs a full OAuth2
-             * resource server (JWT, RFC 8707 audience, RFC 9728 metadata). {@code apikey} accepts a
-             * Stirling per-user API key via the {@code X-API-KEY} header (or {@code Authorization:
-             * Bearer <key>}) and binds the request to that user - the low-friction self-host path,
-             * no external IdP required.
-             */
-            private String mode = "oauth";
-
-            /** OAuth2 issuer URI, e.g. {@code http://localhost:9000}. Required when MCP is on. */
-            private String issuerUri = "";
-
-            /**
-             * JWKS URI. When blank, derived from the issuer's {@code
-             * /.well-known/openid-configuration} document.
-             */
-            private String jwksUri = "";
-
-            /**
-             * RFC 8707 resource identifier of THIS MCP server, e.g. {@code
-             * http://localhost:8080/mcp}. Tokens that do not list this id in their {@code aud}
-             * claim are rejected with HTTP 401.
-             */
-            private String resourceId = "";
-
-            /**
-             * Additional JWT audiences accepted at the MCP endpoint, on top of {@link #resourceId}.
-             * Empty (default) keeps strict RFC 8707 binding. Some IdPs cannot mint
-             * resource-specific audiences - e.g. Supabase's OAuth server always issues {@code
-             * aud=authenticated} - so operators list the audience their IdP actually emits here
-             * (env: {@code MCP_AUTH_ACCEPTEDAUDIENCES}, comma-separated).
-             */
-            private List<String> acceptedAudiences = new ArrayList<>();
-
-            /**
-             * JWT claim whose value is matched against a provisioned Stirling username. Defaults to
-             * {@code sub}; set to {@code email} or {@code preferred_username} to match how your IdP
-             * maps users to Stirling accounts.
-             */
-            private String usernameClaim = "sub";
-
-            /**
-             * When {@code true} (default), a validated token is accepted only if its {@link
-             * #usernameClaim} value resolves to an existing, enabled Stirling user account. Tokens
-             * whose subject has no Stirling account (or a disabled one) are rejected with HTTP 403.
-             * Set to {@code false} only if you intentionally want any IdP-valid token to use MCP
-             * without a local account.
-             */
-            private boolean requireExistingAccount = true;
-        }
-    }
-
    /**
     * Cluster backplane configuration. All keys live under the top-level {@code cluster.*} prefix
     * (e.g. env var {@code CLUSTER_ENABLED}). The master switch is {@link #enabled} and defaults to
@@ -519,14 +376,6 @@ public class ApplicationProperties {
        private String accessibilityStatement;
        private String cookiePolicy;
        private String impressum;
-        private LoginAgreement loginAgreement = new LoginAgreement();
-
-        @Data
-        public static class LoginAgreement {
-            private boolean enabled = false;
-            private boolean showInAnonymousMode = true;
-            private String fallbackText = "";
-        }
    }

    @Data
@@ -595,7 +444,7 @@ public class ApplicationProperties {
        public static class SAML2 {
            private String provider;
            private Boolean enabled = false;
-            private Boolean autoCreateUser = true;
+            private Boolean autoCreateUser = false;
            private Boolean blockRegistration = false;
            private String registrationId = "stirling";

@@ -672,7 +521,7 @@ public class ApplicationProperties {
            private String issuer;
            private String clientId;
            @ToString.Exclude private String clientSecret;
-            private Boolean autoCreateUser = true;
+            private Boolean autoCreateUser = false;
            private Boolean blockRegistration = false;
            private String useAsUsername;
            private Collection<String> scopes = new ArrayList<>();
@@ -743,6 +592,7 @@ public class ApplicationProperties {
        @Data
        public static class Jwt {
            private boolean enableKeystore = true;
+            private boolean enableKeyRotation = false;
            private boolean enableKeyCleanup = true;

            /**
@@ -846,8 +696,8 @@ public class ApplicationProperties {
            @Data
            public static class Trust {
                private boolean serverAsAnchor = true;
-                private boolean useSystemTrust = true;
-                private boolean useMozillaBundle = true;
+                private boolean useSystemTrust = false;
+                private boolean useMozillaBundle = false;
                private boolean useAATL = false;
                private boolean useEUTL = false;
            }
@@ -881,8 +731,8 @@ public class ApplicationProperties {
    public static class System {
        private String defaultLocale;
        private boolean googlevisibility;
-        private boolean showUpdate = true;
-        private boolean showUpdateOnlyAdmin = true;
+        private boolean showUpdate;
+        private boolean showUpdateOnlyAdmin;
        private boolean showSettingsWhenNoLogin = true;
        private boolean customHTMLFiles;
        private String tessdataDir;
@@ -890,10 +740,10 @@ public class ApplicationProperties {
        private Boolean enableAnalytics;
        private Boolean enablePosthog;
        private Boolean enableScarf;
-        private Boolean enableDesktopInstallSlide = true;
+        private Boolean enableDesktopInstallSlide;
        private Datasource datasource;
        private boolean disableSanitize;
-        private int maxDPI = 500;
+        private int maxDPI;
        private boolean enableUrlToPDF;
        private Html html = new Html();
        private CustomPaths customPaths = new CustomPaths();
@@ -907,9 +757,8 @@ public class ApplicationProperties {
        private String frontendUrl; // Frontend URL for invite email links (e.g.

        // 'https://app.example.com'). If not set, falls back to backendUrl.
-        private boolean enableMobileScanner = true; // Enable mobile phone QR code upload feature
+        private boolean enableMobileScanner = false; // Enable mobile phone QR code upload feature
        private MobileScannerSettings mobileScannerSettings = new MobileScannerSettings();
-        private ServerCertificate serverCertificate = new ServerCertificate();

        @Data
        public static class MobileScannerSettings {
@@ -919,16 +768,6 @@ public class ApplicationProperties {
            private boolean stretchToFit = false; // Whether to stretch image to fill page
        }

-        @Data
-        public static class ServerCertificate {
-            private boolean enabled =
-                    true; // Enable server-side "Sign with Stirling-PDF" certificate
-            private String organizationName = "Stirling PDF Inc";
-            private int validity = 365; // Certificate validity in days
-            private boolean regenerateOnStartup =
-                    false; // Generate a new certificate on each startup
-        }
-
        public boolean isAnalyticsEnabled() {
            return this.enableAnalytics != null && this.enableAnalytics;
        }
@@ -1013,7 +852,7 @@ public class ApplicationProperties {
        @Data
        public static class Sharing {
            private boolean enabled = false;
-            private boolean linkEnabled = true;
+            private boolean linkEnabled = false;
            private boolean emailEnabled = false;
            private int linkExpirationDays = 3;
        }
@@ -1028,10 +867,6 @@ public class ApplicationProperties {
        @Data
        public static class Signing {
            private boolean enabled = false;
-
-            // Signing user-picker scope: 'org' (default) = whole instance, anything else =
-            // caller's team only (fail-closed). The saas profile pins 'team'.
-            private String userListScope = "org";
        }
    }

@@ -1187,7 +1022,7 @@ public class ApplicationProperties {

    @Data
    public static class Metrics {
-        private boolean enabled = true;
+        private boolean enabled;
    }

    @Data
@@ -1239,7 +1074,7 @@ public class ApplicationProperties {
        private boolean enableInvites = false;
        private int inviteLinkExpiryHours = 72; // Default: 72 hours (3 days)
        private String host;
-        private int port = 587;
+        private int port;
        private String username;
        @ToString.Exclude private String password;
        private String from;
@@ -1266,10 +1101,10 @@ public class ApplicationProperties {
        @ToString.Exclude private String botToken;
        private String botUsername;
        private String pipelineInboxFolder = "telegram";
-        private Boolean customFolderSuffix = true;
-        private Boolean enableAllowUserIDs = true;
+        private Boolean customFolderSuffix = false;
+        private Boolean enableAllowUserIDs = false;
        private List<Long> allowUserIDs = new ArrayList<>();
-        private Boolean enableAllowChannelIDs = true;
+        private Boolean enableAllowChannelIDs = false;
        private List<Long> allowChannelIDs = new ArrayList<>();
        private long processingTimeoutSeconds = 180;
        private long pollingIntervalMillis = 2000;
@@ -1,6 +1,7 @@
 package stirling.software.common.model;

 import java.nio.file.Path;
+import java.nio.file.Paths;
 import java.time.LocalDateTime;
 import java.time.format.DateTimeFormatter;
 import java.util.Locale;
@@ -23,7 +24,7 @@ public class FileInfo {

    // Converts the file path string to a Path object.
    public Path getFilePathAsPath() {
-        return Path.of(filePath);
+        return Paths.get(filePath);
    }

    // Formats the file size into a human-readable string.
@@ -1,191 +0,0 @@
-package stirling.software.common.pdf;
-
-import java.util.ArrayList;
-import java.util.Collections;
-import java.util.HashMap;
-import java.util.List;
-import java.util.Map;
-
-import stirling.software.jpdfium.text.PageText;
-import stirling.software.jpdfium.text.TextChar;
-import stirling.software.jpdfium.text.TextLine;
-import stirling.software.jpdfium.text.TextWord;
-
-final class HeadingDetector {
-
-    private HeadingDetector() {}
-
-    /** A heading is at most this many words; longer lines are treated as body text. */
-    private static final int MAX_HEADING_WORDS = 12;
-
-    /**
-     * Returns the Markdown heading prefix for a line. The decision combines several signals, never
-     * text matching, so a plain line that merely shares text with a heading is never promoted:
-     *
-     * <ul>
-     *   <li><b>Size</b> — dominant glyph font size vs. the document body median (primary signal).
-     *       Some PDFs encode visual size in the text matrix, so every glyph reports ~1.0; for those
-     *       the line height is used as the proxy instead.
-     *   <li><b>Brevity</b> — headings are short labels; a line over {@value #MAX_HEADING_WORDS}
-     *       words is body text regardless of size.
-     *   <li><b>Not a sentence</b> — a line ending in {@code . ! ?} reads as prose, not a heading.
-     * </ul>
-     *
-     * <p>Boldness is deliberately <em>not</em> a heading signal — a bold-but-not-larger line is
-     * emphasis, not a heading (see {@link #isBoldLabel}); promoting it to {@code #}/{@code ##} is
-     * the main source of false-positive headings.
-     *
-     * <ul>
-     *   <li>size &gt; baseline * 1.4 → {@code "# "}
-     *   <li>size &gt; baseline * 1.2 → {@code "## "}
-     *   <li>otherwise → {@code ""}
-     * </ul>
-     */
-    static String headingPrefix(TextLine line, float medianBodySize, float medianBodyHeight) {
-        String text = line.text().strip();
-        if (text.isEmpty() || wordCount(text) > MAX_HEADING_WORDS || endsLikeSentence(text)) {
-            return "";
-        }
-
-        float dominant = dominantFontSize(line);
-        float value;
-        float baseline;
-        if (dominant > 2f && medianBodySize > 2f) {
-            value = dominant;
-            baseline = medianBodySize;
-        } else {
-            value = line.height();
-            baseline = medianBodyHeight;
-        }
-        if (baseline <= 0f) {
-            return "";
-        }
-
-        float ratio = value / baseline;
-        if (ratio > 1.4f) {
-            return "# ";
-        }
-        if (ratio > 1.2f) {
-            return "## ";
-        }
-        return "";
-    }
-
-    /**
-     * True when a line should be emphasised as bold (rendered {@code **like this**}) rather than
-     * promoted to a heading: it is bold, short, and not a full sentence. Used for bold labels that
-     * are not large enough to be headings.
-     */
-    static boolean isBoldLabel(TextLine line) {
-        String text = line.text().strip();
-        if (text.isEmpty() || wordCount(text) > MAX_HEADING_WORDS || endsLikeSentence(text)) {
-            return false;
-        }
-        return isBold(line);
-    }
-
-    private static int wordCount(String text) {
-        return text.split("\\s+").length;
-    }
-
-    private static boolean endsLikeSentence(String text) {
-        char last = text.charAt(text.length() - 1);
-        return last == '.' || last == '!' || last == '?';
-    }
-
-    /** True when the line's dominant font is bold, inferred from PostScript font names. */
-    private static boolean isBold(TextLine line) {
-        Map<String, Integer> counts = new HashMap<>();
-        for (TextWord word : line.words()) {
-            for (TextChar ch : word.chars()) {
-                if (ch.isWhitespace() || ch.isNewline()) {
-                    continue;
-                }
-                String name = ch.fontName();
-                if (name != null && !name.isBlank()) {
-                    counts.merge(name, 1, Integer::sum);
-                }
-            }
-        }
-        String dominantFont = "";
-        int max = -1;
-        for (Map.Entry<String, Integer> e : counts.entrySet()) {
-            if (e.getValue() > max) {
-                max = e.getValue();
-                dominantFont = e.getKey();
-            }
-        }
-        String lower = dominantFont.toLowerCase(java.util.Locale.ROOT);
-        return lower.contains("bold")
-                || lower.contains("black")
-                || lower.contains("heavy")
-                || lower.contains("semibold");
-    }
-
-    /** Computes the median glyph font size across all pages. */
-    static float medianFontSize(List<PageText> allPages) {
-        List<Float> sizes = new ArrayList<>();
-        for (PageText page : allPages) {
-            for (TextChar ch : page.chars()) {
-                if (!ch.isWhitespace() && !ch.isNewline() && ch.fontSize() > 0f) {
-                    sizes.add(ch.fontSize());
-                }
-            }
-        }
-        return median(sizes, 12f);
-    }
-
-    /** Computes the median TextLine height across all pages. Used when font size is degenerate. */
-    static float medianLineHeight(List<PageText> allPages) {
-        List<Float> heights = new ArrayList<>();
-        for (PageText page : allPages) {
-            for (TextLine line : page.lines()) {
-                if (line.height() > 0f && !line.text().isBlank()) {
-                    heights.add(line.height());
-                }
-            }
-        }
-        return median(heights, 12f);
-    }
-
-    private static float median(List<Float> values, float fallback) {
-        if (values.isEmpty()) {
-            return fallback;
-        }
-        Collections.sort(values);
-        int mid = values.size() / 2;
-        if (values.size() % 2 == 0) {
-            return (values.get(mid - 1) + values.get(mid)) / 2f;
-        }
-        return values.get(mid);
-    }
-
-    /**
-     * Returns the font size that appears most often (by character count) in the given line. Ties
-     * are broken in favour of the larger size.
-     */
-    private static float dominantFontSize(TextLine line) {
-        Map<Float, Integer> counts = new HashMap<>();
-        for (TextWord word : line.words()) {
-            for (TextChar ch : word.chars()) {
-                if (!ch.isWhitespace() && !ch.isNewline() && ch.fontSize() > 0f) {
-                    counts.merge(ch.fontSize(), 1, Integer::sum);
-                }
-            }
-        }
-        if (counts.isEmpty()) {
-            return 0f;
-        }
-        float dominant = 0f;
-        int maxCount = -1;
-        for (Map.Entry<Float, Integer> entry : counts.entrySet()) {
-            int count = entry.getValue();
-            float size = entry.getKey();
-            if (count > maxCount || (count == maxCount && size > dominant)) {
-                maxCount = count;
-                dominant = size;
-            }
-        }
-        return dominant;
-    }
-}
@@ -1,82 +0,0 @@
-package stirling.software.common.pdf;
-
-import stirling.software.jpdfium.text.Table;
-
-final class TableRenderer {
-    private TableRenderer() {}
-
-    /** Renders a Table as a GitHub-Flavoured Markdown table string. */
-    static String render(Table table) {
-        if (table.rowCount() == 0) {
-            return "";
-        }
-
-        String[][] grid = table.asGrid();
-
-        if (table.rowCount() < 2) {
-            // No separator row possible — return plain lines
-            StringBuilder sb = new StringBuilder();
-            for (int c = 0; c < grid[0].length; c++) {
-                if (c > 0) sb.append('\n');
-                sb.append(escape(grid[0][c].trim()));
-            }
-            return sb.toString();
-        }
-
-        int cols = grid[0].length;
-
-        // Compute column widths: max(3, max content length across all rows)
-        int[] widths = new int[cols];
-        for (int c = 0; c < cols; c++) {
-            widths[c] = 3;
-        }
-        for (String[] row : grid) {
-            for (int c = 0; c < cols; c++) {
-                String cell = c < row.length ? row[c].trim() : "";
-                widths[c] = Math.max(widths[c], escape(cell).length());
-            }
-        }
-
-        StringBuilder sb = new StringBuilder();
-
-        // Header row
-        sb.append(buildRow(grid[0], widths, cols));
-        sb.append('\n');
-
-        // Separator row
-        sb.append('|');
-        for (int c = 0; c < cols; c++) {
-            sb.append('-').append("-".repeat(widths[c])).append('-').append('|');
-        }
-        sb.append('\n');
-
-        // Data rows
-        for (int r = 1; r < grid.length; r++) {
-            sb.append(buildRow(grid[r], widths, cols));
-            if (r < grid.length - 1) {
-                sb.append('\n');
-            }
-        }
-
-        return sb.toString();
-    }
-
-    private static String buildRow(String[] row, int[] widths, int cols) {
-        StringBuilder sb = new StringBuilder();
-        sb.append('|');
-        for (int c = 0; c < cols; c++) {
-            String cell = c < row.length ? escape(row[c].trim()) : "";
-            sb.append(' ').append(padRight(cell, widths[c])).append(' ').append('|');
-        }
-        return sb.toString();
-    }
-
-    private static String escape(String cell) {
-        return cell.replace("|", "\\|");
-    }
-
-    private static String padRight(String s, int width) {
-        if (s.length() >= width) return s;
-        return s + " ".repeat(width - s.length());
-    }
-}
@@ -5,7 +5,6 @@ import java.io.IOException;
 import java.io.InputStream;
 import java.io.PipedInputStream;
 import java.io.PipedOutputStream;
-import java.util.Optional;
 import java.util.concurrent.Executors;
 import java.util.concurrent.atomic.AtomicReference;

@@ -18,7 +17,6 @@ import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;

 import stirling.software.common.cluster.FileStore;
-import stirling.software.common.util.JobContext;

 /**
 * Service for storing and retrieving files with unique file IDs. Used by the AutoJobPostMapping
@@ -34,10 +32,8 @@ public class FileStorage {

    private final FileOrUploadService fileOrUploadService;
    private final FileStore fileStore;
-    private final Optional<JobOwnershipService> jobOwnershipService;

    public String storeFile(MultipartFile file) throws IOException {
-        String owner = resolveOwner();
        // Fast path: when Spring buffered the multipart to disk (typical for large uploads), the
        // backing Resource exposes a real File. Hand the Path to the FileStore so it can do a
        // file-to-file copy (Linux sendfile, no copy through Java heap) rather than streaming
@@ -52,7 +48,7 @@ public class FileStorage {
        if (res != null && res.isFile()) {
            try {
                FileStore.Stored stored =
-                        fileStore.store(res.getFile().toPath(), file.getOriginalFilename(), owner);
+                        fileStore.store(res.getFile().toPath(), file.getOriginalFilename());
                log.debug("Stored file with ID: {} (fast path)", stored.fileId());
                return stored.fileId();
            } catch (IOException ex) {
@@ -61,45 +57,40 @@ public class FileStorage {
            }
        }
        try (InputStream in = file.getInputStream()) {
-            FileStore.Stored stored = fileStore.store(in, file.getOriginalFilename(), owner);
+            FileStore.Stored stored = fileStore.store(in, file.getOriginalFilename());
            log.debug("Stored file with ID: {}", stored.fileId());
            return stored.fileId();
        }
    }

    public String storeBytes(byte[] bytes, String originalName) throws IOException {
-        FileStore.Stored stored =
-                fileStore.store(new ByteArrayInputStream(bytes), originalName, resolveOwner());
+        FileStore.Stored stored = fileStore.store(new ByteArrayInputStream(bytes), originalName);
        log.debug("Stored byte array with ID: {}", stored.fileId());
        return stored.fileId();
    }

    public MultipartFile retrieveFile(String fileId) throws IOException {
-        enforceOwnership(fileId);
        byte[] fileData = fileStore.retrieveBytes(fileId);
        return fileOrUploadService.toMockMultipartFile(fileId, fileData);
    }

    public byte[] retrieveBytes(String fileId) throws IOException {
-        enforceOwnership(fileId);
        return fileStore.retrieveBytes(fileId);
    }

    public InputStream retrieveInputStream(String fileId) throws IOException {
-        enforceOwnership(fileId);
        return fileStore.retrieve(fileId);
    }

    public StoredFile storeInputStream(InputStream inputStream, String originalName)
            throws IOException {
-        FileStore.Stored stored = fileStore.store(inputStream, originalName, resolveOwner());
+        FileStore.Stored stored = fileStore.store(inputStream, originalName);
        log.debug("Stored input stream with ID: {}", stored.fileId());
        return new StoredFile(stored.fileId(), stored.size());
    }

    public String storeFromStreamingBody(StreamingResponseBody body, String originalName)
            throws IOException {
-        String owner = resolveOwner();
        // Hold Throwable not IOException: an unchecked failure (NPE, IllegalState, OOM, etc.)
        // from the body writer would otherwise close the pipe with EOF and the consumer would
        // return a truncated file with no error surfaced to the caller.
@@ -124,7 +115,7 @@ public class FileStorage {
                                        }
                                    }
                                });
-                FileStore.Stored stored = fileStore.store(in, originalName, owner);
+                FileStore.Stored stored = fileStore.store(in, originalName);
                Throwable writerErr = bodyError.get();
                if (writerErr != null) {
                    // Body failed mid-write: the FileStore persisted a truncated entry.
@@ -168,62 +159,21 @@ public class FileStorage {

    public String storeFromResource(Resource resource, String originalName) throws IOException {
        try (InputStream in = resource.getInputStream()) {
-            FileStore.Stored stored = fileStore.store(in, originalName, resolveOwner());
+            FileStore.Stored stored = fileStore.store(in, originalName);
            log.debug("Stored Resource with ID: {}", stored.fileId());
            return stored.fileId();
        }
    }

    public boolean deleteFile(String fileId) {
-        enforceOwnership(fileId);
        return fileStore.delete(fileId);
    }

    public boolean fileExists(String fileId) {
-        enforceOwnership(fileId);
        return fileStore.exists(fileId);
    }

    public long getFileSize(String fileId) throws IOException {
-        enforceOwnership(fileId);
        return fileStore.size(fileId);
    }
-
-    private String resolveOwner() {
-        String propagated = JobContext.getOwner();
-        if (propagated != null) {
-            return propagated;
-        }
-        return jobOwnershipService.flatMap(JobOwnershipService::getCurrentUserId).orElse(null);
-    }
-
-    private void enforceOwnership(String fileId) {
-        if (jobOwnershipService.isEmpty()) {
-            return;
-        }
-        Optional<String> currentUser = jobOwnershipService.get().getCurrentUserId();
-        if (currentUser.isEmpty()) {
-            return;
-        }
-        String owner;
-        try {
-            owner = fileStore.getOwner(fileId);
-        } catch (IOException e) {
-            log.warn("Failed to read owner for file {}: {}", fileId, e.getMessage());
-            throw new SecurityException(
-                    "Access denied: could not verify ownership of the requested file");
-        }
-        if (owner == null) {
-            return;
-        }
-        if (!owner.equals(currentUser.get())) {
-            log.warn(
-                    "Access denied: user {} attempted to access file {} owned by {}",
-                    currentUser.get(),
-                    fileId,
-                    owner);
-            throw new SecurityException(
-                    "Access denied: you do not have permission to access this file");
-        }
-    }
 }
@@ -50,16 +50,6 @@ public class InternalApiClient {
                    "^/api/v1/(general|misc|security|convert|filter)(/[A-Za-z0-9_-]+)+$"
                            + "|^/api/v1/ai/tools(/[A-Za-z0-9_-]+)+$");

-    /**
-     * Marker propagated on every internal sub-step dispatch so the saas PAYG interceptor classifies
-     * the call as {@code BillingCategory.AUTOMATION}. By construction every {@link
-     * InternalApiClient#post} caller is an automation surface (pipeline executor, AI workflow,
-     * policy runner) running a child tool inside a parent automation flow — see the saas {@code
-     * PaygChargeInterceptor.determineCategory} precedence chain, where this header dominates any
-     * per-tool {@code @RequiresFeature} annotation.
-     */
-    public static final String AUTOMATION_HEADER = "X-Stirling-Automation";
-
    private final ServletContext servletContext;
    private final UserServiceInterface userService;
    private final TempFileManager tempFileManager;
@@ -106,23 +96,7 @@ public class InternalApiClient {
        if (apiKey != null && !apiKey.isEmpty()) {
            headers.add("X-API-KEY", apiKey);
        }
-        // Tag the sub-step as automation so PAYG bills it under AUTOMATION regardless of which
-        // tool-level @RequiresFeature annotation the dispatched controller carries (e.g. an AI-OCR
-        // step inside a policy run must bill as AUTOMATION, not AI). Set unconditionally because
-        // every caller of this dispatcher is an automation surface by design.
-        headers.add(AUTOMATION_HEADER, "true");

-        // A no-file ai/tools call (e.g. create-pdf-from-html-agent) sends only string params, so
-        // without this RestTemplate would use urlencoded instead of the multipart the controller
-        // expects. File-bearing calls get the right multipart content-type from RestTemplate.
-        boolean isAiTool = endpointPath.startsWith("/api/v1/ai/tools/");
-        boolean hasFilePart =
-                body.values().stream()
-                        .flatMap(java.util.List::stream)
-                        .anyMatch(v -> v instanceof Resource);
-        if (isAiTool && !hasFilePart) {
-            headers.setContentType(MediaType.MULTIPART_FORM_DATA);
-        }
        HttpEntity<MultiValueMap<String, Object>> entity = new HttpEntity<>(body, headers);
        RequestCallback requestCallback = restTemplate.httpEntityCallback(entity, Resource.class);

@@ -89,11 +89,6 @@ public class JobExecutorService {

        String jobId = scopedJobKey;

-        final String jobOwner =
-                jobOwnershipService != null
-                        ? jobOwnershipService.getCurrentUserId().orElse(null)
-                        : null;
-
        long timeoutToUse = customTimeoutMs > 0 ? customTimeoutMs : effectiveTimeoutMs;

        log.debug(
@@ -124,7 +119,6 @@ public class JobExecutorService {
                        try {
                            stirling.software.common.util.JobContext.setJobId(
                                    capturedJobIdForQueue);
-                            stirling.software.common.util.JobContext.setOwner(jobOwner);
                            Object result = work.get();
                            processJobResult(capturedJobIdForQueue, result);
                            return result;
@@ -159,7 +153,6 @@ public class JobExecutorService {
                                    timeoutToUse);

                            stirling.software.common.util.JobContext.setJobId(capturedJobId);
-                            stirling.software.common.util.JobContext.setOwner(jobOwner);
                            Object result = executeWithTimeout(() -> work.get(), timeoutToUse);
                            processJobResult(capturedJobId, result);
                        } catch (TimeoutException te) {
@@ -1,204 +0,0 @@
-package stirling.software.common.service;
-
-import java.io.IOException;
-import java.nio.charset.StandardCharsets;
-import java.nio.file.AtomicMoveNotSupportedException;
-import java.nio.file.Files;
-import java.nio.file.LinkOption;
-import java.nio.file.Path;
-import java.nio.file.StandardCopyOption;
-import java.util.ArrayList;
-import java.util.List;
-import java.util.Set;
-import java.util.TreeSet;
-import java.util.regex.Pattern;
-import java.util.stream.Stream;
-
-import org.springframework.stereotype.Service;
-
-import lombok.extern.slf4j.Slf4j;
-
-import stirling.software.common.configuration.InstallationPathConfig;
-import stirling.software.common.model.ApplicationProperties;
-
-// Resolves login agreement text from customFiles/disclaimer/<locale>.md (read live);
-// enable/visibility come from the legal.loginAgreement settings.
-@Service
-@Slf4j
-public class LoginAgreementService {
-
-    // Locale codes only: rejects path separators and dots so the value can never escape the
-    // disclaimer directory. Matches e.g. en, en-GB, fr-FR, zh-Hant, pt-BR.
-    private static final Pattern LOCALE_PATTERN =
-            Pattern.compile("^[A-Za-z]{2,3}([_-][A-Za-z0-9]{2,8})*$");
-
-    // BCP-47 tags are well under this; the cap also prevents the regex's repetition group
-    // from recursing far enough to overflow the stack on a hostile over-length input.
-    private static final int MAX_LOCALE_LENGTH = 35;
-
-    // Disclaimers are short markdown; cap the read so an oversized file can't be loaded
-    // wholesale into heap on every public request.
-    private static final long MAX_FILE_BYTES = 256 * 1024;
-
-    private final ApplicationProperties applicationProperties;
-
-    public LoginAgreementService(ApplicationProperties applicationProperties) {
-        this.applicationProperties = applicationProperties;
-    }
-
-    public boolean isEnabled() {
-        return config().isEnabled();
-    }
-
-    public boolean isShowInAnonymousMode() {
-        return config().isShowInAnonymousMode();
-    }
-
-    /**
-     * Resolve the markdown to show for the requested language, falling back through the base
-     * language, the configured default locale (and its base), then the configured fallbackText.
-     * Returns an empty string when nothing is configured.
-     */
-    public String resolveContent(String requestedLang) {
-        List<String> candidates = new ArrayList<>();
-        addLocaleCandidates(candidates, requestedLang);
-        addLocaleCandidates(candidates, applicationProperties.getSystem().getDefaultLocale());
-
-        for (String candidate : candidates) {
-            String content = readFileIfExists(candidate);
-            if (content != null && !content.isBlank()) {
-                return content;
-            }
-        }
-
-        String fallback = config().getFallbackText();
-        return fallback == null ? "" : fallback;
-    }
-
-    /**
-     * Admin read of a single locale's raw file. Returns null for an invalid locale, "" if absent.
-     */
-    public String readRawForLocale(String locale) {
-        if (!isValidLocale(locale)) {
-            return null;
-        }
-        String content = readFileIfExists(locale);
-        return content == null ? "" : content;
-    }
-
-    /** Admin write. Blank content deletes the file so it falls back cleanly. */
-    public void writeForLocale(String locale, String content) throws IOException {
-        Path file = resolveLocaleFile(locale);
-        if (file == null) {
-            throw new IllegalArgumentException("Invalid locale: " + locale);
-        }
-        if (content == null || content.isBlank()) {
-            Files.deleteIfExists(file);
-            return;
-        }
-        Files.createDirectories(file.getParent());
-        // Write to a sibling temp file then atomically swap, so a concurrent reader (the public
-        // /login-disclaimer fetch is lockless) never observes a truncated/partial file.
-        Path tmp = Files.createTempFile(file.getParent(), "disclaimer", ".md.tmp");
-        try {
-            Files.writeString(tmp, content, StandardCharsets.UTF_8);
-            try {
-                Files.move(
-                        tmp,
-                        file,
-                        StandardCopyOption.ATOMIC_MOVE,
-                        StandardCopyOption.REPLACE_EXISTING);
-            } catch (AtomicMoveNotSupportedException e) {
-                Files.move(tmp, file, StandardCopyOption.REPLACE_EXISTING);
-            }
-        } finally {
-            Files.deleteIfExists(tmp);
-        }
-    }
-
-    /** Locales that currently have a markdown file, for the admin editor. */
-    public Set<String> listLocalesWithContent() {
-        Set<String> result = new TreeSet<>();
-        Path dir = disclaimerDir();
-        if (!Files.isDirectory(dir)) {
-            return result;
-        }
-        try (Stream<Path> files = Files.list(dir)) {
-            files.filter(Files::isRegularFile)
-                    .map(path -> path.getFileName().toString())
-                    .filter(name -> name.endsWith(".md"))
-                    .map(name -> name.substring(0, name.length() - ".md".length()))
-                    .filter(this::isValidLocale)
-                    .forEach(result::add);
-        } catch (IOException e) {
-            log.warn("Failed listing login agreement files", e);
-        }
-        return result;
-    }
-
-    private ApplicationProperties.Legal.LoginAgreement config() {
-        return applicationProperties.getLegal().getLoginAgreement();
-    }
-
-    private Path disclaimerDir() {
-        return Path.of(InstallationPathConfig.getCustomFilesPath(), "disclaimer").normalize();
-    }
-
-    private void addLocaleCandidates(List<String> out, String locale) {
-        if (!isValidLocale(locale)) {
-            return;
-        }
-        if (!out.contains(locale)) {
-            out.add(locale);
-        }
-        String base = locale.split("[_-]", 2)[0];
-        if (!base.equals(locale) && !out.contains(base)) {
-            out.add(base);
-        }
-    }
-
-    private String readFileIfExists(String locale) {
-        Path file = resolveLocaleFile(locale);
-        if (file == null) {
-            return null;
-        }
-        try {
-            // NOFOLLOW_LINKS: a symlinked entry is treated as non-regular and skipped, so a
-            // planted symlink can't expose files outside the disclaimer dir via the public read.
-            if (Files.isRegularFile(file, LinkOption.NOFOLLOW_LINKS)) {
-                if (Files.size(file) > MAX_FILE_BYTES) {
-                    log.warn(
-                            "Login agreement file for locale {} exceeds {} bytes; ignoring",
-                            locale,
-                            MAX_FILE_BYTES);
-                    return null;
-                }
-                return Files.readString(file, StandardCharsets.UTF_8);
-            }
-        } catch (IOException e) {
-            log.warn("Failed reading login agreement file for locale {}", locale, e);
-        }
-        return null;
-    }
-
-    private Path resolveLocaleFile(String locale) {
-        if (!isValidLocale(locale)) {
-            return null;
-        }
-        Path dir = disclaimerDir();
-        Path file = dir.resolve(locale + ".md").normalize();
-        // Defence in depth: the regex already blocks separators, but confirm containment.
-        if (!file.startsWith(dir)) {
-            return null;
-        }
-        return file;
-    }
-
-    private boolean isValidLocale(String locale) {
-        // Length check BEFORE the regex: LOCALE_PATTERN's repetition group recurses one stack
-        // frame per repeat in java.util.regex, so an unbounded input could overflow the stack.
-        return locale != null
-                && locale.length() <= MAX_LOCALE_LENGTH
-                && LOCALE_PATTERN.matcher(locale).matches();
-    }
-}
@@ -3,6 +3,7 @@ package stirling.software.common.service;
 import java.io.IOException;
 import java.nio.file.Files;
 import java.nio.file.Path;
+import java.nio.file.Paths;
 import java.util.ArrayList;
 import java.util.HashMap;
 import java.util.List;
@@ -34,7 +35,7 @@ public class MobileScannerService {
    public MobileScannerService() throws IOException {
        // Create temp directory for mobile scanner uploads
        this.tempDirectory =
-                Path.of(System.getProperty("java.io.tmpdir"), "stirling-mobile-scanner");
+                Paths.get(System.getProperty("java.io.tmpdir"), "stirling-mobile-scanner");
        Files.createDirectories(tempDirectory);
        log.info("Mobile scanner temp directory: {}", tempDirectory);
    }
@@ -7,7 +7,6 @@ import java.time.format.DateTimeFormatter;
 import java.util.Calendar;

 import org.apache.pdfbox.pdmodel.PDDocument;
-import org.apache.pdfbox.pdmodel.PDDocumentInformation;
 import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.beans.factory.annotation.Qualifier;
 import org.springframework.stereotype.Service;
@@ -18,9 +17,6 @@ import stirling.software.common.model.PdfMetadata;
@Service
 public class PdfMetadataService {

-    /** ({@code {labels}}). Written by the classify-and-label tool. */
-    public static final String CLASSIFICATION_KEY = "StirlingPDFClassification";
-
    private final ApplicationProperties applicationProperties;
    private final String stirlingPDFLabel;
    private final UserServiceInterface userService;
@@ -181,14 +177,4 @@ public class PdfMetadataService {
        }
        pdf.getDocumentInformation().setAuthor(author);
    }
-
-    /**
-     * Write the document classifier's JSON result into the custom Info-dictionary field {@link
-     * #CLASSIFICATION_KEY}, leaving all other metadata untouched.
-     */
-    public void setClassificationMetadata(PDDocument pdf, String classificationJson) {
-        PDDocumentInformation info = pdf.getDocumentInformation();
-        info.setCustomMetadataValue(CLASSIFICATION_KEY, classificationJson);
-        pdf.setDocumentInformation(info);
-    }
 }
@@ -8,7 +8,7 @@ import java.lang.management.OperatingSystemMXBean;
 import java.lang.management.RuntimeMXBean;
 import java.lang.management.ThreadMXBean;
 import java.nio.file.Files;
-import java.nio.file.Path;
+import java.nio.file.Paths;
 import java.util.HashMap;
 import java.util.Locale;
 import java.util.Map;
@@ -160,7 +160,7 @@ public class PostHogService {
    }

    private boolean isRunningInDocker() {
-        return Files.exists(Path.of("/.dockerenv"));
+        return Files.exists(Paths.get("/.dockerenv"));
    }

    private Map<String, Object> getDockerMetrics() {
@@ -1,21 +1,11 @@
 package stirling.software.common.service;

-import java.util.List;
-
 /** Provides metadata about tool endpoints for internal dispatch. */
 public interface ToolMetadataService {

    /** Returns true if the given operation path accepts multiple input files. */
    boolean isMultiInput(String operationPath);

-    /**
-     * Returns the file extensions (lowercase, no leading dot, e.g. {@code "pdf"}) that the
-     * operation accepts as input ({@code output=false}) or produces as output ({@code
-     * output=true}), derived from the endpoint's declared type. Returns {@code null} when the
-     * endpoint declares no specific type, which callers should treat as "any type accepted".
-     */
-    List<String> getExtensionTypes(boolean output, String operationPath);
-
    /**
     * Returns true when the endpoint's ZIP response is a transport for multiple typed results and
     * should be unpacked: multi-output endpoints (Type:SIMO / Type:MIMO) and wrapper declarations
@@ -255,7 +255,7 @@ public class GeneralUtils {
        String pattern = locationPattern;
        if (pattern.startsWith("file:")) {
            String rawPath = pattern.substring(5).replace("\\*", "").replace("/*", "");
-            Path normalizePath = Path.of(rawPath).normalize();
+            Path normalizePath = Paths.get(rawPath).normalize();
            pattern = "file:" + normalizePath.toString().replace("\\", "/") + "/*";
        }
        return ResourcePatternUtils.getResourcePatternResolver(resourceLoader)
@@ -837,7 +837,7 @@ public class GeneralUtils {
    }

    public boolean createDir(String path) {
-        Path folder = Path.of(path);
+        Path folder = Paths.get(path);
        if (!Files.exists(folder)) {
            try {
                Files.createDirectories(folder);
@@ -867,7 +867,7 @@ public class GeneralUtils {

    public void saveKeyToSettings(String key, Object newValue) throws IOException {
        String[] keyArray = key.split("\\.");
-        Path settingsPath = Path.of(InstallationPathConfig.getSettingsPath());
+        Path settingsPath = Paths.get(InstallationPathConfig.getSettingsPath());
        YamlHelper settingsYaml = new YamlHelper(settingsPath);
        settingsYaml.updateValue(Arrays.asList(keyArray), newValue);
        settingsYaml.saveOverride(settingsPath);
@@ -888,7 +888,7 @@ public class GeneralUtils {
            return;
        }

-        Path settingsPath = Path.of(InstallationPathConfig.getSettingsPath());
+        Path settingsPath = Paths.get(InstallationPathConfig.getSettingsPath());
        YamlHelper settingsYaml = new YamlHelper(settingsPath);

        // Apply all updates to the same YamlHelper instance
@@ -974,11 +974,11 @@ public class GeneralUtils {
     */
    public void extractPipeline() throws IOException {
        Path pipelineDir =
-                Path.of(InstallationPathConfig.getPipelinePath(), DEFAULT_WEBUI_CONFIGS_DIR);
+                Paths.get(InstallationPathConfig.getPipelinePath(), DEFAULT_WEBUI_CONFIGS_DIR);
        Files.createDirectories(pipelineDir);

        for (String name : DEFAULT_VALID_PIPELINE) {
-            if (!Path.of(name).getFileName().toString().equals(name)) {
+            if (!Paths.get(name).getFileName().toString().equals(name)) {
                log.error("Invalid pipeline file name: {}", name);
                throw new IllegalArgumentException("Invalid pipeline file name: " + name);
            }
@@ -1014,7 +1014,7 @@ public class GeneralUtils {
            throw new IllegalArgumentException(
                    "scriptName must not contain path traversal characters");
        }
-        if (!Path.of(scriptName).getFileName().toString().equals(scriptName)) {
+        if (!Paths.get(scriptName).getFileName().toString().equals(scriptName)) {
            throw new IllegalArgumentException(
                    "scriptName must not contain path traversal characters");
        }
@@ -1024,7 +1024,7 @@ public class GeneralUtils {
                    "scriptName must be either 'png_to_webp.py' or 'split_photos.py'");
        }

-        Path scriptsDir = Path.of(InstallationPathConfig.getScriptsPath(), PYTHON_SCRIPTS_DIR);
+        Path scriptsDir = Paths.get(InstallationPathConfig.getScriptsPath(), PYTHON_SCRIPTS_DIR);
        Files.createDirectories(scriptsDir);

        Path target = scriptsDir.resolve(scriptName);
@@ -1185,181 +1185,23 @@ public class GeneralUtils {
    }

    public String getLocalNetworkIp() {
-        String routed = detectLocalIpViaDefaultRoute();
-        if (routed != null) {
-            return routed;
-        }
        try {
-            return selectBestSiteLocalIp(collectInterfaceInfo());
-        } catch (Exception e) {
-            log.warn("Failed to detect local network IP", e);
-            return null;
-        }
-    }
-
-    private String detectLocalIpViaDefaultRoute() {
-        try (DatagramSocket socket = new DatagramSocket()) {
-            socket.connect(InetAddress.getByName("8.8.8.8"), 53);
-            InetAddress local = socket.getLocalAddress();
-            if (local instanceof Inet4Address
-                    && !local.isAnyLocalAddress()
-                    && !local.isLoopbackAddress()
-                    && !local.isLinkLocalAddress()) {
-                return local.getHostAddress();
+            Enumeration<NetworkInterface> interfaces = NetworkInterface.getNetworkInterfaces();
+            if (interfaces == null) return null;
+            while (interfaces.hasMoreElements()) {
+                NetworkInterface iface = interfaces.nextElement();
+                if (!iface.isUp() || iface.isLoopback() || iface.isVirtual()) continue;
+                Enumeration<InetAddress> addresses = iface.getInetAddresses();
+                while (addresses.hasMoreElements()) {
+                    InetAddress addr = addresses.nextElement();
+                    if (addr instanceof Inet4Address && addr.isSiteLocalAddress()) {
+                        return addr.getHostAddress();
+                    }
+                }
            }
        } catch (Exception e) {
-            log.debug("Default-route IP detection failed; will scan interfaces", e);
+            log.warn("Failed to detect local network IP", e);
        }
        return null;
    }
-
-    private List<NetworkInterfaceInfo> collectInterfaceInfo() throws SocketException {
-        List<NetworkInterfaceInfo> infos = new ArrayList<>();
-        Enumeration<NetworkInterface> interfaces = NetworkInterface.getNetworkInterfaces();
-        if (interfaces == null) {
-            return infos;
-        }
-        while (interfaces.hasMoreElements()) {
-            NetworkInterface iface = interfaces.nextElement();
-
-            List<String> siteLocalIpv4s = new ArrayList<>();
-            Enumeration<InetAddress> addresses = iface.getInetAddresses();
-            while (addresses.hasMoreElements()) {
-                InetAddress addr = addresses.nextElement();
-                if (addr instanceof Inet4Address && addr.isSiteLocalAddress()) {
-                    siteLocalIpv4s.add(addr.getHostAddress());
-                }
-            }
-            if (siteLocalIpv4s.isEmpty()) {
-                continue;
-            }
-
-            try {
-                byte[] mac = iface.getHardwareAddress();
-                infos.add(
-                        new NetworkInterfaceInfo(
-                                iface.getName(),
-                                iface.getDisplayName(),
-                                iface.getIndex(),
-                                iface.isUp(),
-                                iface.isLoopback(),
-                                iface.isPointToPoint(),
-                                iface.isVirtual(),
-                                mac != null && mac.length > 0,
-                                siteLocalIpv4s));
-            } catch (SocketException e) {
-                log.debug("Skipping interface {} while scanning for local IP", iface.getName(), e);
-            }
-        }
-        return infos;
-    }
-
-    static String selectBestSiteLocalIp(List<NetworkInterfaceInfo> interfaces) {
-        return interfaces.stream()
-                .filter(i -> i.up() && !i.loopback() && !i.pointToPoint() && !i.virtual())
-                .filter(i -> !isLikelyVirtualInterface(i.name(), i.displayName()))
-                .flatMap(
-                        i ->
-                                i.siteLocalIpv4s().stream()
-                                        .map(
-                                                ip ->
-                                                        new ScoredAddress(
-                                                                ip,
-                                                                scoreInterface(i, ip),
-                                                                i.index())))
-                .max(
-                        Comparator.comparingInt(ScoredAddress::score)
-                                .thenComparing(
-                                        Comparator.comparingInt(ScoredAddress::interfaceIndex)
-                                                .reversed()))
-                .map(ScoredAddress::ip)
-                .orElse(null);
-    }
-
-    private static int scoreInterface(NetworkInterfaceInfo iface, String ip) {
-        int score = 0;
-        if (isLikelyPhysicalInterface(iface.name(), iface.displayName())) {
-            score += 100;
-        }
-        if (iface.hasHardwareAddress()) {
-            score += 20;
-        }
-        if (ip.startsWith("192.168.")) {
-            score += 30;
-        } else if (ip.startsWith("10.")) {
-            score += 20;
-        } else {
-            score += 5;
-        }
-        return score;
-    }
-
-    static boolean isLikelyVirtualInterface(String name, String displayName) {
-        String n = name == null ? "" : name.toLowerCase(Locale.ROOT);
-        String d = displayName == null ? "" : displayName.toLowerCase(Locale.ROOT);
-        String[] namePrefixes = {
-            "tun", "tap", "utun", "veth", "virbr", "vmnet", "docker", "br-", "wg", "ppp", "awdl",
-            "llw"
-        };
-        for (String prefix : namePrefixes) {
-            if (n.startsWith(prefix)) {
-                return true;
-            }
-        }
-        String[] displayMarkers = {
-            "vmware",
-            "virtualbox",
-            "virtual box",
-            "vbox",
-            "hyper-v",
-            "hyperv",
-            "vethernet",
-            "windows subsystem for linux",
-            "wsl",
-            "docker",
-            "tap-windows",
-            "tunnel",
-            "vpn",
-            "zerotier",
-            "tailscale",
-            "bluetooth",
-            "teredo",
-            "isatap",
-            "loopback",
-            "pseudo",
-            "virtual"
-        };
-        for (String marker : displayMarkers) {
-            if (d.contains(marker)) {
-                return true;
-            }
-        }
-        return false;
-    }
-
-    private static boolean isLikelyPhysicalInterface(String name, String displayName) {
-        String n = name == null ? "" : name.toLowerCase(Locale.ROOT);
-        String d = displayName == null ? "" : displayName.toLowerCase(Locale.ROOT);
-        return n.startsWith("eth")
-                || n.startsWith("en")
-                || n.startsWith("wl")
-                || n.startsWith("em")
-                || d.contains("ethernet")
-                || d.contains("wi-fi")
-                || d.contains("wifi")
-                || d.contains("wireless");
-    }
-
-    record NetworkInterfaceInfo(
-            String name,
-            String displayName,
-            int index,
-            boolean up,
-            boolean loopback,
-            boolean pointToPoint,
-            boolean virtual,
-            boolean hasHardwareAddress,
-            List<String> siteLocalIpv4s) {}
-
-    private record ScoredAddress(String ip, int score, int interfaceIndex) {}
 }
@@ -4,6 +4,7 @@ import java.io.File;
 import java.net.URISyntaxException;
 import java.nio.file.Files;
 import java.nio.file.Path;
+import java.nio.file.Paths;

 import lombok.extern.slf4j.Slf4j;

@@ -19,7 +20,7 @@ public class JarPathUtil {
    public static Path currentJar() {
        try {
            Path jar =
-                    Path.of(
+                    Paths.get(
                                    JarPathUtil.class
                                            .getProtectionDomain()
                                            .getCodeSource()
@@ -60,14 +61,14 @@ public class JarPathUtil {
        }

        // Location 2: ./build/libs/ (development build)
-        possibleLocations[1] = Path.of("build", "libs", "restart-helper.jar").toAbsolutePath();
+        possibleLocations[1] = Paths.get("build", "libs", "restart-helper.jar").toAbsolutePath();

        // Location 3: app/common/build/libs/ (multi-module build)
        possibleLocations[2] =
-                Path.of("app", "common", "build", "libs", "restart-helper.jar").toAbsolutePath();
+                Paths.get("app", "common", "build", "libs", "restart-helper.jar").toAbsolutePath();

        // Location 4: Current working directory
-        possibleLocations[3] = Path.of("restart-helper.jar").toAbsolutePath();
+        possibleLocations[3] = Paths.get("restart-helper.jar").toAbsolutePath();

        // Check each location
        for (Path location : possibleLocations) {
@@ -1,9 +1,8 @@
 package stirling.software.common.util;

-/** Thread-local context for passing job ID and owner across async boundaries */
+/** Thread-local context for passing job ID across async boundaries */
 public class JobContext {
    private static final ThreadLocal<String> CURRENT_JOB_ID = new ThreadLocal<>();
-    private static final ThreadLocal<String> CURRENT_OWNER = new ThreadLocal<>();

    public static void setJobId(String jobId) {
        CURRENT_JOB_ID.set(jobId);
@@ -13,16 +12,7 @@ public class JobContext {
        return CURRENT_JOB_ID.get();
    }

-    public static void setOwner(String owner) {
-        CURRENT_OWNER.set(owner);
-    }
-
-    public static String getOwner() {
-        return CURRENT_OWNER.get();
-    }
-
    public static void clear() {
        CURRENT_JOB_ID.remove();
-        CURRENT_OWNER.remove();
    }
 }
@@ -1,310 +0,0 @@
-package stirling.software.common.util;
-
-import java.io.ByteArrayInputStream;
-import java.io.ByteArrayOutputStream;
-import java.io.IOException;
-import java.util.ArrayList;
-import java.util.List;
-import java.util.Locale;
-import java.util.Set;
-import java.util.zip.ZipEntry;
-import java.util.zip.ZipInputStream;
-import java.util.zip.ZipOutputStream;
-
-import javax.xml.XMLConstants;
-import javax.xml.parsers.DocumentBuilder;
-import javax.xml.parsers.DocumentBuilderFactory;
-import javax.xml.parsers.ParserConfigurationException;
-import javax.xml.transform.OutputKeys;
-import javax.xml.transform.Transformer;
-import javax.xml.transform.TransformerException;
-import javax.xml.transform.TransformerFactory;
-import javax.xml.transform.dom.DOMSource;
-import javax.xml.transform.stream.StreamResult;
-
-import org.springframework.stereotype.Component;
-import org.w3c.dom.Document;
-import org.w3c.dom.Element;
-import org.w3c.dom.NamedNodeMap;
-import org.w3c.dom.Node;
-import org.w3c.dom.NodeList;
-import org.xml.sax.SAXException;
-
-import io.github.pixee.security.ZipSecurity;
-
-import lombok.extern.slf4j.Slf4j;
-
-import stirling.software.common.model.ApplicationProperties;
-import stirling.software.common.service.SsrfProtectionService;
-
-// Strips external refs from OOXML/ODF uploads so LibreOffice can't be made to fetch them.
-@Component
-@Slf4j
-public class OfficeDocumentSanitizer {
-
-    private static final Set<String> OOXML_EXTENSIONS =
-            Set.of(
-                    "docx", "docm", "dotx", "dotm", "xlsx", "xlsm", "xltx", "xltm", "pptx", "pptm",
-                    "potx", "potm", "ppsx", "ppsm");
-
-    private static final Set<String> ODF_EXTENSIONS =
-            Set.of(
-                    "odt", "ott", "ods", "ots", "odp", "otp", "odg", "otg", "odf", "odc", "odi",
-                    "odm");
-
-    private static final Set<String> ODF_XML_PARTS =
-            Set.of("content.xml", "styles.xml", "meta.xml", "settings.xml");
-
-    private final SsrfProtectionService ssrfProtectionService;
-    private final ApplicationProperties applicationProperties;
-
-    public OfficeDocumentSanitizer(
-            SsrfProtectionService ssrfProtectionService,
-            ApplicationProperties applicationProperties) {
-        this.ssrfProtectionService = ssrfProtectionService;
-        this.applicationProperties = applicationProperties;
-    }
-
-    public boolean isSanitizableExtension(String extension) {
-        if (extension == null) {
-            return false;
-        }
-        String lower = extension.toLowerCase(Locale.ROOT);
-        return OOXML_EXTENSIONS.contains(lower) || ODF_EXTENSIONS.contains(lower);
-    }
-
-    public byte[] sanitize(byte[] documentBytes, String extension) throws IOException {
-        if (documentBytes == null || documentBytes.length == 0) {
-            throw new IOException("Office document input is empty or null");
-        }
-        if (applicationProperties.getSystem().isDisableSanitize()) {
-            log.debug("Office document sanitization disabled by configuration");
-            return documentBytes;
-        }
-        if (!isSanitizableExtension(extension)) {
-            return documentBytes;
-        }
-
-        ByteArrayOutputStream out = new ByteArrayOutputStream(documentBytes.length);
-        try (ZipInputStream zipIn =
-                        ZipSecurity.createHardenedInputStream(
-                                new ByteArrayInputStream(documentBytes));
-                ZipOutputStream zipOut = new ZipOutputStream(out)) {
-
-            ZipEntry entry;
-            while ((entry = zipIn.getNextEntry()) != null) {
-                String name = entry.getName();
-                byte[] bytes = entry.isDirectory() ? new byte[0] : zipIn.readAllBytes();
-
-                if (!entry.isDirectory()) {
-                    bytes = sanitizeEntry(name, bytes);
-                }
-
-                ZipEntry outEntry = new ZipEntry(name);
-                if (entry.getComment() != null) {
-                    outEntry.setComment(entry.getComment());
-                }
-                if (entry.getExtra() != null) {
-                    outEntry.setExtra(entry.getExtra());
-                }
-                zipOut.putNextEntry(outEntry);
-                if (!entry.isDirectory()) {
-                    zipOut.write(bytes);
-                }
-                zipOut.closeEntry();
-            }
-        }
-        return out.toByteArray();
-    }
-
-    private byte[] sanitizeEntry(String entryName, byte[] entryBytes) {
-        String lower = entryName.toLowerCase(Locale.ROOT);
-        try {
-            if (lower.endsWith(".rels")) {
-                return sanitizeOoxmlRels(entryBytes);
-            }
-            if (isOdfXmlPart(lower)) {
-                return sanitizeOdfXml(entryBytes);
-            }
-        } catch (ParserConfigurationException
-                | SAXException
-                | IOException
-                | TransformerException e) {
-            log.warn(
-                    "Failed to parse XML part '{}' for sanitization, leaving as-is: {}",
-                    entryName,
-                    e.getMessage());
-        }
-        return entryBytes;
-    }
-
-    private boolean isOdfXmlPart(String lowerName) {
-        int slash = lowerName.lastIndexOf('/');
-        String base = slash >= 0 ? lowerName.substring(slash + 1) : lowerName;
-        return ODF_XML_PARTS.contains(base);
-    }
-
-    private byte[] sanitizeOoxmlRels(byte[] xmlBytes)
-            throws IOException, ParserConfigurationException, SAXException, TransformerException {
-        Document doc = parseSecurely(xmlBytes);
-        Element root = doc.getDocumentElement();
-        if (root == null) {
-            return xmlBytes;
-        }
-        NodeList relationships = root.getElementsByTagNameNS("*", "Relationship");
-        List<Node> toRemove = new ArrayList<>();
-        for (int i = 0; i < relationships.getLength(); i++) {
-            Node node = relationships.item(i);
-            NamedNodeMap attrs = node.getAttributes();
-            if (attrs == null) {
-                continue;
-            }
-            Node targetMode = attrs.getNamedItem("TargetMode");
-            if (targetMode == null || !"external".equalsIgnoreCase(targetMode.getNodeValue())) {
-                continue;
-            }
-            Node target = attrs.getNamedItem("Target");
-            String targetValue = target == null ? "" : target.getNodeValue();
-            if (isAdminAllowed(targetValue)) {
-                continue;
-            }
-            log.warn(
-                    "Stripping OOXML external relationship target: {}",
-                    truncateForLog(targetValue));
-            toRemove.add(node);
-        }
-        if (toRemove.isEmpty()) {
-            return xmlBytes;
-        }
-        for (Node n : toRemove) {
-            n.getParentNode().removeChild(n);
-        }
-        return serializeDocument(doc);
-    }
-
-    private byte[] sanitizeOdfXml(byte[] xmlBytes)
-            throws IOException, ParserConfigurationException, SAXException, TransformerException {
-        Document doc = parseSecurely(xmlBytes);
-        Element root = doc.getDocumentElement();
-        if (root == null) {
-            return xmlBytes;
-        }
-        boolean modified = stripExternalHrefs(root);
-        if (!modified) {
-            return xmlBytes;
-        }
-        return serializeDocument(doc);
-    }
-
-    private boolean stripExternalHrefs(Node node) {
-        boolean modified = false;
-        if (node.getNodeType() == Node.ELEMENT_NODE) {
-            NamedNodeMap attrs = node.getAttributes();
-            List<String> hrefAttrsToRemove = new ArrayList<>();
-            for (int i = 0; i < attrs.getLength(); i++) {
-                Node attr = attrs.item(i);
-                String name = attr.getNodeName();
-                if (name == null) {
-                    continue;
-                }
-                String lower = name.toLowerCase(Locale.ROOT);
-                if (!(lower.equals("xlink:href")
-                        || lower.endsWith(":href")
-                        || lower.equals("href"))) {
-                    continue;
-                }
-                String value = attr.getNodeValue();
-                if (!isExternalUrl(value)) {
-                    continue;
-                }
-                if (isAdminAllowed(value)) {
-                    continue;
-                }
-                log.warn(
-                        "Stripping ODF external href attribute ({}): {}",
-                        name,
-                        truncateForLog(value));
-                hrefAttrsToRemove.add(name);
-            }
-            Element element = (Element) node;
-            for (String attrName : hrefAttrsToRemove) {
-                element.removeAttribute(attrName);
-                modified = true;
-            }
-        }
-        NodeList children = node.getChildNodes();
-        for (int i = 0; i < children.getLength(); i++) {
-            if (stripExternalHrefs(children.item(i))) {
-                modified = true;
-            }
-        }
-        return modified;
-    }
-
-    private boolean isExternalUrl(String url) {
-        if (url == null) {
-            return false;
-        }
-        String trimmed = url.trim().toLowerCase(Locale.ROOT);
-        if (trimmed.isEmpty() || trimmed.startsWith("#") || trimmed.startsWith("../")) {
-            return false;
-        }
-        return trimmed.startsWith("http://")
-                || trimmed.startsWith("https://")
-                || trimmed.startsWith("ftp://")
-                || trimmed.startsWith("ftps://")
-                || trimmed.startsWith("file:")
-                || trimmed.startsWith("smb:")
-                || trimmed.startsWith("\\\\")
-                || trimmed.startsWith("//");
-    }
-
-    // Preserved only with an explicit allowedDomains entry; MEDIUM default would admit public URLs.
-    private boolean isAdminAllowed(String url) {
-        if (ssrfProtectionService == null || url == null || url.isBlank()) {
-            return false;
-        }
-        ApplicationProperties.Html.UrlSecurity config =
-                applicationProperties.getSystem().getHtml().getUrlSecurity();
-        if (config == null
-                || config.getAllowedDomains() == null
-                || config.getAllowedDomains().isEmpty()) {
-            return false;
-        }
-        return ssrfProtectionService.isUrlAllowed(url);
-    }
-
-    private Document parseSecurely(byte[] xmlBytes)
-            throws ParserConfigurationException, SAXException, IOException {
-        DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance();
-        factory.setFeature(XMLConstants.FEATURE_SECURE_PROCESSING, true);
-        factory.setFeature("http://apache.org/xml/features/disallow-doctype-decl", true);
-        factory.setFeature("http://xml.org/sax/features/external-general-entities", false);
-        factory.setFeature("http://xml.org/sax/features/external-parameter-entities", false);
-        factory.setFeature("http://apache.org/xml/features/nonvalidating/load-external-dtd", false);
-        factory.setXIncludeAware(false);
-        factory.setExpandEntityReferences(false);
-        factory.setNamespaceAware(true);
-        DocumentBuilder builder = factory.newDocumentBuilder();
-        return builder.parse(new ByteArrayInputStream(xmlBytes));
-    }
-
-    private byte[] serializeDocument(Document doc) throws TransformerException {
-        TransformerFactory tf = TransformerFactory.newInstance();
-        tf.setFeature(XMLConstants.FEATURE_SECURE_PROCESSING, true);
-        Transformer transformer = tf.newTransformer();
-        transformer.setOutputProperty(OutputKeys.ENCODING, "UTF-8");
-        transformer.setOutputProperty(OutputKeys.INDENT, "no");
-        transformer.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION, "no");
-        ByteArrayOutputStream baos = new ByteArrayOutputStream();
-        transformer.transform(new DOMSource(doc), new StreamResult(baos));
-        return baos.toByteArray();
-    }
-
-    private String truncateForLog(String value) {
-        if (value == null) {
-            return "null";
-        }
-        return value.length() > 80 ? value.substring(0, 80) + "..." : value;
-    }
-}
@@ -57,16 +57,6 @@ public class RequestUriUtils {
            return true;
        }

-        // Admin portal SPA shell (mounted at /processor — must match the frontend
-        // PORTAL_BASENAME). Served publicly like the editor root so a direct nav /
-        // refresh to /processor loads the app (the JWT lives in localStorage, not a
-        // cookie, so the server can't authenticate the navigation itself). The
-        // portal gates access via its own auth gate + RequirePortalAccess, and its
-        // data APIs stay protected, so serving the shell pre-auth is safe.
-        if (normalizedUri.equals("/processor") || normalizedUri.startsWith("/processor/")) {
-            return true;
-        }
-
        // Treat common static file extensions as static resources
        return normalizedUri.endsWith(".svg")
                || normalizedUri.endsWith(".png")
@@ -244,7 +244,10 @@ public class SvgSanitizer {
            return false;
        }

-        return true;
+        return normalized.startsWith("http://")
+                || normalized.startsWith("https://")
+                || normalized.startsWith("//")
+                || normalized.startsWith("file:");
    }

    private boolean isUrlAllowed(String url) {
@@ -1,481 +0,0 @@
-package stirling.software.SPDF.config;
-
-import static org.junit.jupiter.api.Assertions.assertEquals;
-import static org.junit.jupiter.api.Assertions.assertFalse;
-import static org.junit.jupiter.api.Assertions.assertNotNull;
-import static org.junit.jupiter.api.Assertions.assertNull;
-import static org.junit.jupiter.api.Assertions.assertTrue;
-
-import java.util.List;
-import java.util.Set;
-
-import org.junit.jupiter.api.BeforeEach;
-import org.junit.jupiter.api.DisplayName;
-import org.junit.jupiter.api.Nested;
-import org.junit.jupiter.api.Test;
-
-import stirling.software.SPDF.config.EndpointConfiguration.DisableReason;
-import stirling.software.SPDF.config.EndpointConfiguration.EndpointAvailability;
-import stirling.software.common.model.ApplicationProperties;
-
-/**
- * Unit tests for {@link EndpointConfiguration}. The class wires up its endpoint/group registry in
- * {@code init()} during construction and then applies environment overrides. We build it with a
- * real {@link ApplicationProperties} (whose System/Endpoints sub-objects are non-null by default)
- * so the constructor runs cleanly without any mocking.
- */
-class EndpointConfigurationGapTest {
-
-    private ApplicationProperties applicationProperties;
-
-    /**
-     * Construct an EndpointConfiguration with the given pro flag and current applicationProperties.
-     */
-    private EndpointConfiguration build(boolean runningProOrHigher) {
-        return new EndpointConfiguration(applicationProperties, runningProOrHigher);
-    }
-
-    /** Default config: not pro, no removals, url-to-pdf disabled (default System flag is false). */
-    private EndpointConfiguration buildDefault() {
-        return build(false);
-    }
-
-    @BeforeEach
-    void setUp() {
-        applicationProperties = new ApplicationProperties();
-    }
-
-    @Nested
-    @DisplayName("endpointKeyForUri (static)")
-    class EndpointKeyForUriTests {
-
-        @Test
-        @DisplayName("returns null for null uri")
-        void nullUri() {
-            assertNull(EndpointConfiguration.endpointKeyForUri(null));
-        }
-
-        @Test
-        @DisplayName("returns null when uri does not contain /api/v1")
-        void notApiPath() {
-            assertNull(EndpointConfiguration.endpointKeyForUri("/foo/bar/baz"));
-            assertNull(EndpointConfiguration.endpointKeyForUri("https://example.com/home"));
-        }
-
-        @Test
-        @DisplayName("returns null when uri has too few path segments")
-        void tooFewSegments() {
-            // "/api/v1/general" splits to ["", "api", "v1", "general"] -> length 4, not > 4
-            assertNull(EndpointConfiguration.endpointKeyForUri("/api/v1/general"));
-        }
-
-        @Test
-        @DisplayName("extracts plain endpoint key from a standard /api/v1/<group>/<endpoint> uri")
-        void plainEndpoint() {
-            assertEquals(
-                    "remove-pages",
-                    EndpointConfiguration.endpointKeyForUri("/api/v1/general/remove-pages"));
-        }
-
-        @Test
-        @DisplayName("builds a <from>-to-<to> key for convert endpoints")
-        void convertEndpoint() {
-            assertEquals(
-                    "pdf-to-img",
-                    EndpointConfiguration.endpointKeyForUri("/api/v1/convert/pdf/img"));
-        }
-
-        @Test
-        @DisplayName("convert path without a target segment falls back to the segment after group")
-        void convertWithoutTarget() {
-            // "/api/v1/convert/pdf" -> length 5, the convert branch needs length > 5
-            assertEquals("pdf", EndpointConfiguration.endpointKeyForUri("/api/v1/convert/pdf"));
-        }
-    }
-
-    @Nested
-    @DisplayName("enable / disable endpoint")
-    class EnableDisableEndpointTests {
-
-        @Test
-        @DisplayName("a freshly registered endpoint is enabled by default")
-        void enabledByDefault() {
-            EndpointConfiguration config = buildDefault();
-            assertTrue(config.isEndpointEnabled("merge-pdfs"));
-        }
-
-        @Test
-        @DisplayName("disableEndpoint marks the endpoint disabled")
-        void disableEndpoint() {
-            EndpointConfiguration config = buildDefault();
-            config.disableEndpoint("merge-pdfs");
-            assertFalse(config.isEndpointEnabled("merge-pdfs"));
-        }
-
-        @Test
-        @DisplayName("enableEndpoint re-enables a previously disabled endpoint")
-        void reEnableEndpoint() {
-            EndpointConfiguration config = buildDefault();
-            config.disableEndpoint("merge-pdfs");
-            assertFalse(config.isEndpointEnabled("merge-pdfs"));
-            config.enableEndpoint("merge-pdfs");
-            assertTrue(config.isEndpointEnabled("merge-pdfs"));
-        }
-
-        @Test
-        @DisplayName("leading slash is normalized away on disable")
-        void leadingSlashNormalizedOnDisable() {
-            EndpointConfiguration config = buildDefault();
-            config.disableEndpoint("/merge-pdfs");
-            // both forms resolve to the same key
-            assertFalse(config.isEndpointEnabled("merge-pdfs"));
-            assertFalse(config.isEndpointEnabled("/merge-pdfs"));
-        }
-
-        @Test
-        @DisplayName("isEndpointEnabled tolerates a leading slash on the query")
-        void leadingSlashOnQuery() {
-            EndpointConfiguration config = buildDefault();
-            assertTrue(config.isEndpointEnabled("/merge-pdfs"));
-        }
-
-        @Test
-        @DisplayName("disabling clears with enable, removing the disable reason")
-        void enableClearsReason() {
-            EndpointConfiguration config = buildDefault();
-            config.disableEndpoint("split-pages", DisableReason.DEPENDENCY);
-            assertEquals(
-                    DisableReason.DEPENDENCY,
-                    config.getEndpointAvailability("split-pages").getReason());
-            config.enableEndpoint("split-pages");
-            EndpointAvailability availability = config.getEndpointAvailability("split-pages");
-            assertTrue(availability.isEnabled());
-            assertNull(availability.getReason());
-        }
-    }
-
-    @Nested
-    @DisplayName("isEndpointEnabledForUri")
-    class IsEndpointEnabledForUriTests {
-
-        @Test
-        @DisplayName("translates a /api/v1 uri to a key and reports its status")
-        void translatesUri() {
-            EndpointConfiguration config = buildDefault();
-            assertTrue(config.isEndpointEnabledForUri("/api/v1/general/merge-pdfs"));
-            config.disableEndpoint("merge-pdfs");
-            assertFalse(config.isEndpointEnabledForUri("/api/v1/general/merge-pdfs"));
-        }
-
-        @Test
-        @DisplayName("falls back to treating a non-api uri as a raw key")
-        void fallsBackToRawKey() {
-            EndpointConfiguration config = buildDefault();
-            config.disableEndpoint("merge-pdfs");
-            // non-api path: key resolution returns null, so the uri itself is used as the key
-            assertFalse(config.isEndpointEnabledForUri("merge-pdfs"));
-        }
-    }
-
-    @Nested
-    @DisplayName("group enable / disable")
-    class GroupTests {
-
-        @Test
-        @DisplayName("a functional group with all endpoints enabled reports enabled")
-        void functionalGroupEnabled() {
-            EndpointConfiguration config = buildDefault();
-            assertTrue(config.isGroupEnabled("PageOps"));
-        }
-
-        @Test
-        @DisplayName("disabling a functional group cascades to all its endpoints")
-        void disableFunctionalGroupCascades() {
-            EndpointConfiguration config = buildDefault();
-            config.disableGroup("PageOps");
-            assertFalse(config.isGroupEnabled("PageOps"));
-            assertFalse(config.isEndpointEnabled("remove-pages"));
-            assertFalse(config.isEndpointEnabled("split-pages"));
-        }
-
-        @Test
-        @DisplayName("re-enabling a functional group re-enables its endpoints")
-        void enableFunctionalGroupRestores() {
-            EndpointConfiguration config = buildDefault();
-            config.disableGroup("PageOps");
-            assertFalse(config.isEndpointEnabled("remove-pages"));
-            config.enableGroup("PageOps");
-            assertTrue(config.isEndpointEnabled("remove-pages"));
-            assertTrue(config.isGroupEnabled("PageOps"));
-        }
-
-        @Test
-        @DisplayName("a functional group with one disabled endpoint is not enabled")
-        void functionalGroupWithDisabledEndpoint() {
-            EndpointConfiguration config = buildDefault();
-            config.disableEndpoint("remove-pages");
-            assertFalse(config.isGroupEnabled("PageOps"));
-        }
-
-        @Test
-        @DisplayName("disabledGroups reflects disabled groups and getDisabledGroups returns a copy")
-        void getDisabledGroupsReturnsCopy() {
-            EndpointConfiguration config = buildDefault();
-            config.disableGroup("PageOps");
-            Set<String> disabled = config.getDisabledGroups();
-            assertTrue(disabled.contains("PageOps"));
-            // mutating the returned set must not affect internal state
-            disabled.clear();
-            assertTrue(config.getDisabledGroups().contains("PageOps"));
-        }
-
-        @Test
-        @DisplayName("an unknown group with no endpoints is not enabled")
-        void unknownGroupNotEnabled() {
-            EndpointConfiguration config = buildDefault();
-            assertFalse(config.isGroupEnabled("NoSuchGroupXyz"));
-        }
-    }
-
-    @Nested
-    @DisplayName("tool group semantics")
-    class ToolGroupTests {
-
-        @Test
-        @DisplayName("a tool group is enabled until explicitly disabled")
-        void toolGroupEnabledUntilDisabled() {
-            EndpointConfiguration config = buildDefault();
-            assertTrue(config.isGroupEnabled("qpdf"));
-            config.disableGroup("qpdf");
-            assertFalse(config.isGroupEnabled("qpdf"));
-        }
-
-        @Test
-        @DisplayName("disabling a tool group does NOT cascade to its endpoints directly")
-        void toolGroupNoCascade() {
-            EndpointConfiguration config = buildDefault();
-            // repair has alternatives (qpdf, Ghostscript); disabling only qpdf keeps it enabled
-            config.disableGroup("qpdf");
-            assertTrue(config.isEndpointEnabled("repair"));
-        }
-
-        @Test
-        @DisplayName("endpoint with alternatives is disabled only when all tool groups are gone")
-        void allAlternativesDisabled() {
-            EndpointConfiguration config = buildDefault();
-            config.disableGroup("qpdf");
-            config.disableGroup("Ghostscript");
-            // repair's only alternatives are qpdf and Ghostscript
-            assertFalse(config.isEndpointEnabled("repair"));
-        }
-
-        @Test
-        @DisplayName("endpoint with a still-enabled alternative stays enabled")
-        void oneAlternativeRemains() {
-            EndpointConfiguration config = buildDefault();
-            // compress-pdf alternatives: qpdf, Ghostscript, Java
-            config.disableGroup("qpdf");
-            config.disableGroup("Ghostscript");
-            assertTrue(config.isEndpointEnabled("compress-pdf"));
-            config.disableGroup("Java");
-            assertFalse(config.isEndpointEnabled("compress-pdf"));
-        }
-
-        @Test
-        @DisplayName("single-dependency endpoint (no alternatives) disabled when its tool group is")
-        void singleDependencyDisabled() {
-            EndpointConfiguration config = buildDefault();
-            // pdf-to-epub depends on Calibre, no alternatives registered
-            assertTrue(config.isEndpointEnabled("pdf-to-epub"));
-            config.disableGroup("Calibre");
-            assertFalse(config.isEndpointEnabled("pdf-to-epub"));
-        }
-    }
-
-    @Nested
-    @DisplayName("addEndpointToGroup / addEndpointAlternative")
-    class RegistrationTests {
-
-        @Test
-        @DisplayName("addEndpointToGroup makes the endpoint part of the group")
-        void addEndpointToGroup() {
-            EndpointConfiguration config = buildDefault();
-            config.addEndpointToGroup("CustomGroup", "custom-endpoint");
-            Set<String> endpoints = config.getEndpointsForGroup("CustomGroup");
-            assertTrue(endpoints.contains("custom-endpoint"));
-        }
-
-        @Test
-        @DisplayName("disabling a custom functional group disables its added endpoint")
-        void customFunctionalGroupCascades() {
-            EndpointConfiguration config = buildDefault();
-            config.addEndpointToGroup("CustomGroup", "custom-endpoint");
-            assertTrue(config.isEndpointEnabled("custom-endpoint"));
-            config.disableGroup("CustomGroup");
-            assertFalse(config.isEndpointEnabled("custom-endpoint"));
-        }
-
-        @Test
-        @DisplayName("getEndpointsForGroup returns an empty set for unknown groups")
-        void unknownGroupEmptySet() {
-            EndpointConfiguration config = buildDefault();
-            Set<String> endpoints = config.getEndpointsForGroup("NoSuchGroupXyz");
-            assertNotNull(endpoints);
-            assertTrue(endpoints.isEmpty());
-        }
-    }
-
-    @Nested
-    @DisplayName("getEndpointAvailability / determineDisableReason")
-    class AvailabilityTests {
-
-        @Test
-        @DisplayName("an enabled endpoint has a null disable reason")
-        void enabledHasNullReason() {
-            EndpointConfiguration config = buildDefault();
-            EndpointAvailability availability = config.getEndpointAvailability("merge-pdfs");
-            assertTrue(availability.isEnabled());
-            assertNull(availability.getReason());
-        }
-
-        @Test
-        @DisplayName("explicit disable preserves the supplied reason")
-        void explicitDisableReason() {
-            EndpointConfiguration config = buildDefault();
-            config.disableEndpoint("merge-pdfs", DisableReason.DEPENDENCY);
-            EndpointAvailability availability = config.getEndpointAvailability("merge-pdfs");
-            assertFalse(availability.isEnabled());
-            assertEquals(DisableReason.DEPENDENCY, availability.getReason());
-        }
-
-        @Test
-        @DisplayName("default disableEndpoint reason is CONFIG")
-        void defaultDisableReasonIsConfig() {
-            EndpointConfiguration config = buildDefault();
-            config.disableEndpoint("merge-pdfs");
-            assertEquals(
-                    DisableReason.CONFIG, config.getEndpointAvailability("merge-pdfs").getReason());
-        }
-
-        @Test
-        @DisplayName("endpoint disabled via functional group reports the group's reason")
-        void functionalGroupReason() {
-            EndpointConfiguration config = buildDefault();
-            config.disableGroup("PageOps", DisableReason.DEPENDENCY);
-            EndpointAvailability availability = config.getEndpointAvailability("crop");
-            assertFalse(availability.isEnabled());
-            // crop is disabled both via group cascade and group membership; reason is DEPENDENCY
-            assertEquals(DisableReason.DEPENDENCY, availability.getReason());
-        }
-    }
-
-    @Nested
-    @DisplayName("getAllEndpoints")
-    class GetAllEndpointsTests {
-
-        @Test
-        @DisplayName("aggregates endpoints across all groups")
-        void aggregatesAcrossGroups() {
-            EndpointConfiguration config = buildDefault();
-            Set<String> all = config.getAllEndpoints();
-            assertTrue(all.contains("merge-pdfs"));
-            assertTrue(all.contains("compress-pdf"));
-            assertTrue(all.contains("ocr-pdf"));
-            assertFalse(all.isEmpty());
-        }
-
-        @Test
-        @DisplayName("custom endpoints registered after init appear in getAllEndpoints")
-        void includesCustomEndpoints() {
-            EndpointConfiguration config = buildDefault();
-            config.addEndpointToGroup("CustomGroup", "brand-new-endpoint");
-            assertTrue(config.getAllEndpoints().contains("brand-new-endpoint"));
-        }
-    }
-
-    @Nested
-    @DisplayName("environment / constructor driven configuration")
-    class EnvironmentConfigTests {
-
-        @Test
-        @DisplayName("url-to-pdf is disabled when enableUrlToPDF is false (default)")
-        void urlToPdfDisabledByDefault() {
-            EndpointConfiguration config = buildDefault();
-            assertFalse(config.isEndpointEnabled("url-to-pdf"));
-        }
-
-        @Test
-        @DisplayName("url-to-pdf stays enabled when enableUrlToPDF is true")
-        void urlToPdfEnabledWhenFlagSet() {
-            applicationProperties.getSystem().setEnableUrlToPDF(true);
-            EndpointConfiguration config = build(false);
-            assertTrue(config.isEndpointEnabled("url-to-pdf"));
-        }
-
-        @Test
-        @DisplayName("endpoints.toRemove disables the listed endpoints at construction")
-        void endpointsToRemove() {
-            applicationProperties
-                    .getEndpoints()
-                    .setToRemove(List.of(" merge-pdfs ", "split-pages"));
-            EndpointConfiguration config = build(false);
-            // values are trimmed before disabling
-            assertFalse(config.isEndpointEnabled("merge-pdfs"));
-            assertFalse(config.isEndpointEnabled("split-pages"));
-        }
-
-        @Test
-        @DisplayName("endpoints.groupsToRemove disables the listed groups at construction")
-        void groupsToRemove() {
-            applicationProperties.getEndpoints().setGroupsToRemove(List.of(" PageOps "));
-            EndpointConfiguration config = build(false);
-            assertTrue(config.getDisabledGroups().contains("PageOps"));
-            assertFalse(config.isEndpointEnabled("remove-pages"));
-        }
-
-        @Test
-        @DisplayName("non-pro build disables the enterprise group")
-        void nonProDisablesEnterprise() {
-            EndpointConfiguration config = build(false);
-            assertTrue(config.getDisabledGroups().contains("enterprise"));
-        }
-
-        @Test
-        @DisplayName("pro build does not disable the enterprise group")
-        void proDoesNotDisableEnterprise() {
-            EndpointConfiguration config = build(true);
-            assertFalse(config.getDisabledGroups().contains("enterprise"));
-        }
-    }
-
-    @Nested
-    @DisplayName("getEndpointStatuses (Lombok getter) and logging summary")
-    class MiscTests {
-
-        @Test
-        @DisplayName("getEndpointStatuses reflects explicit disable state")
-        void endpointStatusesReflectDisable() {
-            EndpointConfiguration config = buildDefault();
-            config.disableEndpoint("merge-pdfs");
-            assertEquals(Boolean.FALSE, config.getEndpointStatuses().get("merge-pdfs"));
-        }
-
-        @Test
-        @DisplayName("logDisabledEndpointsSummary runs without throwing")
-        void logSummaryDoesNotThrow() {
-            EndpointConfiguration config = buildDefault();
-            config.disableGroup("PageOps");
-            config.disableGroup("qpdf");
-            // purely a smoke test of the logging branch coverage
-            config.logDisabledEndpointsSummary();
-        }
-
-        @Test
-        @DisplayName("logDisabledEndpointsSummary runs when nothing is disabled")
-        void logSummaryNothingDisabled() {
-            applicationProperties.getSystem().setEnableUrlToPDF(true);
-            EndpointConfiguration config = build(true);
-            config.logDisabledEndpointsSummary();
-        }
-    }
-}
@@ -0,0 +1,153 @@
+package stirling.software.SPDF.pdf.parser;
+
+import static org.assertj.core.api.Assertions.assertThat;
+import static stirling.software.SPDF.pdf.parser.PdfModels.*;
+
+import java.util.List;
+
+import org.junit.jupiter.api.Test;
+
+/**
+ * Unit tests for {@link LineAlignmentTableParser}, focused on the coincident-line merge logic and
+ * column-grid construction.
+ */
+class LineAlignmentTableParserTest {
+
+    private final LineAlignmentTableParser parser = new LineAlignmentTableParser();
+
+    // ── mergeCoincidentLines ─────────────────────────────────────────────────────────────────────
+
+    @Test
+    void mergeCoincidentLines_singleLine_unchanged() {
+        var lines = List.of(tokenized(rawLine(10f, 100f, "Revenue")));
+        assertThat(parser.mergeCoincidentLines(lines)).hasSize(1);
+    }
+
+    @Test
+    void mergeCoincidentLines_distinctYLines_unchanged() {
+        // Two lines at different y positions — must NOT be merged.
+        var lines =
+                List.of(
+                        tokenized(rawLine(10f, 100f, "Revenue")),
+                        tokenized(rawLine(10f, 115f, "Cost")));
+        assertThat(parser.mergeCoincidentLines(lines)).hasSize(2);
+    }
+
+    @Test
+    void mergeCoincidentLines_sameY_merged() {
+        // Simulates a financial-table row split by LineBuilder at the column gap:
+        //   label fragment at x=72  → "Revenue"
+        //   value fragment at x=350 → "1,234"
+        // Both have y=100. After merge they should form one TokenizedLine.
+        var label = rawLine(72f, 100f, "Revenue");
+        var value = rawLine(350f, 100f, "1,234");
+
+        var merged = parser.mergeCoincidentLines(List.of(tokenized(label), tokenized(value)));
+
+        assertThat(merged).hasSize(1);
+        // The merged line should contain tokens from both halves.
+        var tokens = merged.get(0).all();
+        assertThat(tokens.stream().map(t -> t.text()).toList())
+                .containsExactlyInAnyOrder("Revenue", "1,234");
+    }
+
+    @Test
+    void mergeCoincidentLines_sameY_mergedLineHasCorrectBounds() {
+        var label = rawLine(72f, 100f, "Revenue"); // 7 chars × 6pt = 42pt wide → right = 114
+        var value = rawLine(350f, 100f, "1,234"); // 5 chars × 6pt = 30pt wide → right = 380
+
+        var merged = parser.mergeCoincidentLines(List.of(tokenized(label), tokenized(value)));
+
+        var bounds = merged.get(0).line().bounds();
+        assertThat(bounds.x()).isEqualTo(72f);
+        assertThat(bounds.right()).isEqualTo(380f);
+    }
+
+    @Test
+    void mergeCoincidentLines_withinTolerance_merged() {
+        // Lines 1.5pt apart (within ROW_MERGE_TOLERANCE_PT = 2pt) should merge.
+        var a = rawLine(10f, 100.0f, "Alpha");
+        var b = rawLine(200f, 101.5f, "99");
+
+        var merged = parser.mergeCoincidentLines(List.of(tokenized(a), tokenized(b)));
+        assertThat(merged).hasSize(1);
+    }
+
+    @Test
+    void mergeCoincidentLines_beyondTolerance_notMerged() {
+        // Lines 3pt apart (beyond ROW_MERGE_TOLERANCE_PT = 2pt) should NOT merge.
+        var a = rawLine(10f, 100.0f, "Alpha");
+        var b = rawLine(200f, 103.0f, "99");
+
+        var merged = parser.mergeCoincidentLines(List.of(tokenized(a), tokenized(b)));
+        assertThat(merged).hasSize(2);
+    }
+
+    @Test
+    void mergeCoincidentLines_threeCoincident_allMerged() {
+        // Three fragments at the same y (e.g. wide financial table with two value columns).
+        var a = rawLine(72f, 100f, "Revenue");
+        var b = rawLine(300f, 100f, "1,234");
+        var c = rawLine(400f, 100f, "5,678");
+
+        var merged = parser.mergeCoincidentLines(List.of(tokenized(a), tokenized(b), tokenized(c)));
+        assertThat(merged).hasSize(1);
+        assertThat(merged.get(0).all()).hasSize(3);
+    }
+
+    @Test
+    void mergeCoincidentLines_coincidentPairFollowedByDistinctLine_twoGroups() {
+        var a = rawLine(72f, 100f, "Revenue");
+        var b = rawLine(350f, 100f, "1,234"); // same y as a → merges with a
+        var c = rawLine(10f, 115f, "Expenses"); // different y → stays separate
+
+        var merged = parser.mergeCoincidentLines(List.of(tokenized(a), tokenized(b), tokenized(c)));
+        assertThat(merged).hasSize(2);
+    }
+
+    @Test
+    void mergeCoincidentLines_numericAnchorStatus_correctAfterMerge() {
+        // After merging, the combined line should be an anchor (≥2 numeric tokens).
+        // "Revenue" alone → not an anchor. "1,234  567" alone → anchor.
+        // Merged → anchor with at least 2 numerics.
+        var label = rawLine(72f, 100f, "Revenue");
+        var values = rawLineMultiWord(350f, 100f, "1,234", 30f, "567", 30f);
+
+        var merged = parser.mergeCoincidentLines(List.of(tokenized(label), tokenized(values)));
+
+        assertThat(merged).hasSize(1);
+        assertThat(merged.get(0).isAnchor()).isTrue();
+    }
+
+    // ── helpers ──────────────────────────────────────────────────────────────────────────────────
+
+    /** Creates a RawLine with a single TextFragment of the given text at the given position. */
+    private static RawLine rawLine(float x, float y, String text) {
+        float width = text.length() * 6f; // ~6pt per char — rough but consistent
+        float height = 12f;
+        Bounds bounds = new Bounds(x, y, width, height);
+        TextFragment fragment =
+                new TextFragment("tf-test", text, bounds, y + height, 11f, "Helvetica", false);
+        return new RawLine("ln-test", List.of(fragment), bounds, 1);
+    }
+
+    /**
+     * Creates a RawLine with two TextFragments representing two words separated by a small gap.
+     * Used to simulate a values-only line with multiple numeric tokens.
+     */
+    private static RawLine rawLineMultiWord(
+            float x, float y, String word1, float w1, String word2, float w2) {
+        float height = 12f;
+        Bounds b1 = new Bounds(x, y, w1, height);
+        Bounds b2 = new Bounds(x + w1 + 5f, y, w2, height);
+        TextFragment f1 = new TextFragment("tf-1", word1, b1, y + height, 11f, "Helvetica", false);
+        TextFragment f2 = new TextFragment("tf-2", word2, b2, y + height, 11f, "Helvetica", false);
+        Bounds lineBounds = new Bounds(x, y, x + w1 + 5f + w2 - x, height);
+        return new RawLine("ln-test", List.of(f1, f2), lineBounds, 1);
+    }
+
+    /** Tokenises a RawLine via the parser's own tokenise logic (package-private access). */
+    private LineAlignmentTableParser.TokenizedLine tokenized(RawLine line) {
+        return parser.tokenize(line);
+    }
+}
@@ -1,191 +0,0 @@
-package stirling.software.SPDF.pdf.parser;
-
-import static org.assertj.core.api.Assertions.assertThat;
-import static org.assertj.core.api.Assertions.within;
-
-import java.awt.geom.Point2D;
-import java.awt.image.BufferedImage;
-import java.io.ByteArrayOutputStream;
-import java.util.List;
-
-import org.apache.pdfbox.Loader;
-import org.apache.pdfbox.cos.COSName;
-import org.apache.pdfbox.pdmodel.PDDocument;
-import org.apache.pdfbox.pdmodel.PDPage;
-import org.apache.pdfbox.pdmodel.PDPageContentStream;
-import org.apache.pdfbox.pdmodel.common.PDRectangle;
-import org.apache.pdfbox.pdmodel.graphics.image.LosslessFactory;
-import org.apache.pdfbox.pdmodel.graphics.image.PDImageXObject;
-import org.junit.jupiter.api.DisplayName;
-import org.junit.jupiter.api.Nested;
-import org.junit.jupiter.api.Test;
-
-import stirling.software.SPDF.pdf.parser.PageImageLocator.ImageBox;
-
-/**
- * Unit tests for {@link PageImageLocator}. PDFs are built in memory with PDFBox so each test is
- * deterministic and needs no fixtures or native libraries. The locator transforms the image unit
- * square through the CTM, so an image drawn at {@code (x, y)} with size {@code (w, h)} must yield
- * the box {@code (x, y, x+w, y+h)}.
- */
-class PageImageLocatorTest {
-
-    /** A tiny opaque raster; pixel content is irrelevant, only its placement matters. */
-    private static PDImageXObject tinyImage(PDDocument doc) throws Exception {
-        BufferedImage img = new BufferedImage(4, 4, BufferedImage.TYPE_INT_RGB);
-        return LosslessFactory.createFromImage(doc, img);
-    }
-
-    /** Builds a one-page PDF that draws one image at the given placement. */
-    private static byte[] pdfWithImageAt(float x, float y, float w, float h) throws Exception {
-        try (PDDocument doc = new PDDocument()) {
-            PDPage page = new PDPage(PDRectangle.A4);
-            doc.addPage(page);
-            PDImageXObject image = tinyImage(doc);
-            try (PDPageContentStream cs = new PDPageContentStream(doc, page)) {
-                cs.drawImage(image, x, y, w, h);
-            }
-            return save(doc);
-        }
-    }
-
-    private static byte[] save(PDDocument doc) throws Exception {
-        ByteArrayOutputStream baos = new ByteArrayOutputStream();
-        doc.save(baos);
-        return baos.toByteArray();
-    }
-
-    @Nested
-    @DisplayName("drawImage bounding boxes")
-    class DrawImageBoxes {
-
-        @Test
-        @DisplayName("a single image yields one box with the page index and CTM-derived bounds")
-        void singleImageBox() throws Exception {
-            byte[] pdf = pdfWithImageAt(100f, 200f, 50f, 80f);
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                PageImageLocator locator = new PageImageLocator(doc.getPage(0), 0);
-                locator.processPage(doc.getPage(0));
-
-                List<ImageBox> boxes = locator.getImageBoxes();
-                assertThat(boxes).hasSize(1);
-                ImageBox box = boxes.get(0);
-                assertThat(box.pageIndex()).isZero();
-                assertThat(box.x1()).isCloseTo(100f, within(0.5f));
-                assertThat(box.y1()).isCloseTo(200f, within(0.5f));
-                assertThat(box.x2()).isCloseTo(150f, within(0.5f));
-                assertThat(box.y2()).isCloseTo(280f, within(0.5f));
-            }
-        }
-
-        @Test
-        @DisplayName("the supplied page index is stored on every box")
-        void pageIndexStored() throws Exception {
-            byte[] pdf = pdfWithImageAt(10f, 10f, 20f, 20f);
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                PageImageLocator locator = new PageImageLocator(doc.getPage(0), 7);
-                locator.processPage(doc.getPage(0));
-                assertThat(locator.getImageBoxes().get(0).pageIndex()).isEqualTo(7);
-            }
-        }
-
-        @Test
-        @DisplayName("two images on one page yield two boxes")
-        void twoImages() throws Exception {
-            try (PDDocument doc = new PDDocument()) {
-                PDPage page = new PDPage(PDRectangle.A4);
-                doc.addPage(page);
-                PDImageXObject image = tinyImage(doc);
-                try (PDPageContentStream cs = new PDPageContentStream(doc, page)) {
-                    cs.drawImage(image, 50f, 50f, 30f, 30f);
-                    cs.drawImage(image, 200f, 400f, 60f, 40f);
-                }
-                byte[] pdf = save(doc);
-                try (PDDocument reopened = Loader.loadPDF(pdf)) {
-                    PageImageLocator locator = new PageImageLocator(reopened.getPage(0), 0);
-                    locator.processPage(reopened.getPage(0));
-                    assertThat(locator.getImageBoxes()).hasSize(2);
-                }
-            }
-        }
-
-        @Test
-        @DisplayName("a page with no images yields no boxes")
-        void noImages() throws Exception {
-            try (PDDocument doc = new PDDocument()) {
-                doc.addPage(new PDPage(PDRectangle.A4));
-                byte[] pdf = save(doc);
-                try (PDDocument reopened = Loader.loadPDF(pdf)) {
-                    PageImageLocator locator = new PageImageLocator(reopened.getPage(0), 0);
-                    locator.processPage(reopened.getPage(0));
-                    assertThat(locator.getImageBoxes()).isEmpty();
-                }
-            }
-        }
-
-        @Test
-        @DisplayName("getImageBoxes is empty before any page is processed")
-        void emptyBeforeProcessing() throws Exception {
-            try (PDDocument doc = new PDDocument()) {
-                doc.addPage(new PDPage(PDRectangle.A4));
-                PageImageLocator locator = new PageImageLocator(doc.getPage(0), 0);
-                assertThat(locator.getImageBoxes()).isEmpty();
-            }
-        }
-    }
-
-    @Nested
-    @DisplayName("path operation no-ops")
-    class PathNoOps {
-
-        private PageImageLocator newLocator() {
-            PDPage page = new PDPage(PDRectangle.A4);
-            return new PageImageLocator(page, 0);
-        }
-
-        @Test
-        @DisplayName("moveTo updates the current point")
-        void moveToUpdatesPoint() {
-            PageImageLocator locator = newLocator();
-            locator.moveTo(12f, 34f);
-            Point2D current = locator.getCurrentPoint();
-            assertThat(current.getX()).isEqualTo(12d);
-            assertThat(current.getY()).isEqualTo(34d);
-        }
-
-        @Test
-        @DisplayName("lineTo updates the current point")
-        void lineToUpdatesPoint() {
-            PageImageLocator locator = newLocator();
-            locator.lineTo(5f, 6f);
-            assertThat(locator.getCurrentPoint().getX()).isEqualTo(5d);
-            assertThat(locator.getCurrentPoint().getY()).isEqualTo(6d);
-        }
-
-        @Test
-        @DisplayName("curveTo updates the current point to the final control point")
-        void curveToUpdatesPoint() {
-            PageImageLocator locator = newLocator();
-            locator.curveTo(1f, 1f, 2f, 2f, 9f, 8f);
-            assertThat(locator.getCurrentPoint().getX()).isEqualTo(9d);
-            assertThat(locator.getCurrentPoint().getY()).isEqualTo(8d);
-        }
-
-        @Test
-        @DisplayName("rectangle, clip, path and shading operations are no-ops that do not throw")
-        void otherOpsDoNotThrow() {
-            PageImageLocator locator = newLocator();
-            Point2D p = new Point2D.Float(0f, 0f);
-            // None of these record anything or alter state; they must simply not throw.
-            locator.appendRectangle(p, p, p, p);
-            locator.clip(0);
-            locator.closePath();
-            locator.endPath();
-            locator.strokePath();
-            locator.fillPath(0);
-            locator.fillAndStrokePath(0);
-            locator.shadingFill(COSName.getPDFName("Sh0"));
-            assertThat(locator.getImageBoxes()).isEmpty();
-        }
-    }
-}
@@ -1,345 +0,0 @@
-package stirling.software.SPDF.pdf.parser;
-
-import static org.junit.jupiter.api.Assertions.assertEquals;
-import static org.junit.jupiter.api.Assertions.assertFalse;
-import static org.junit.jupiter.api.Assertions.assertNotNull;
-import static org.junit.jupiter.api.Assertions.assertSame;
-import static org.junit.jupiter.api.Assertions.assertTrue;
-import static stirling.software.SPDF.pdf.parser.PdfModels.RawPage;
-import static stirling.software.SPDF.pdf.parser.PdfModels.TableCell;
-import static stirling.software.SPDF.pdf.parser.PdfModels.TableFragment;
-import static stirling.software.SPDF.pdf.parser.PdfModels.TableRow;
-
-import java.awt.Color;
-import java.io.ByteArrayOutputStream;
-import java.util.List;
-
-import org.apache.pdfbox.Loader;
-import org.apache.pdfbox.pdmodel.PDDocument;
-import org.apache.pdfbox.pdmodel.PDPage;
-import org.apache.pdfbox.pdmodel.PDPageContentStream;
-import org.apache.pdfbox.pdmodel.common.PDRectangle;
-import org.apache.pdfbox.pdmodel.font.PDType1Font;
-import org.apache.pdfbox.pdmodel.font.Standard14Fonts;
-import org.junit.jupiter.api.DisplayName;
-import org.junit.jupiter.api.Nested;
-import org.junit.jupiter.api.Test;
-
-/**
- * Unit tests for {@link TabulaTableParser}. Tables are built in-memory with PDFBox so the tests are
- * deterministic and need no fixtures, network, or external processes.
- */
-class TabulaTableParserGapTest {
-
-    private final TabulaTableParser parser = new TabulaTableParser();
-
-    // ── error / empty branches ───────────────────────────────────────────────
-
-    @Nested
-    @DisplayName("Empty and error branches")
-    class EmptyAndErrorBranches {
-
-        @Test
-        @DisplayName("page number 0 is out of Tabula's 1-based range -> empty list, no throw")
-        void pageNumberZeroReturnsEmpty() throws Exception {
-            byte[] pdf = pdfWithText(new String[] {"hello"});
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                List<TableFragment> result = parser.parse(doc, 0);
-                assertNotNull(result);
-                assertTrue(result.isEmpty());
-            }
-        }
-
-        @Test
-        @DisplayName("page number beyond the document -> empty list, exception swallowed")
-        void pageNumberOutOfRangeReturnsEmpty() throws Exception {
-            byte[] pdf = pdfWithText(new String[] {"hello"});
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                List<TableFragment> result = parser.parse(doc, 99);
-                assertNotNull(result);
-                assertTrue(result.isEmpty());
-            }
-        }
-
-        @Test
-        @DisplayName("negative page number -> empty list")
-        void negativePageNumberReturnsEmpty() throws Exception {
-            byte[] pdf = pdfWithText(new String[] {"hello"});
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                assertTrue(parser.parse(doc, -5).isEmpty());
-            }
-        }
-
-        @Test
-        @DisplayName("lattice mode on a page with no ruled lines -> no tables")
-        void latticeWithNoRulingsReturnsEmpty() throws Exception {
-            byte[] pdf = pdfWithText(new String[] {"just some prose", "no table here"});
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                List<TableFragment> result = parser.parse(doc, new RawPage(1, 0f, 0f, List.of()));
-                assertNotNull(result);
-                assertTrue(
-                        result.isEmpty(), "borderless text must not be detected in lattice mode");
-            }
-        }
-
-        @Test
-        @DisplayName("blank page in lattice mode -> empty list")
-        void blankPageLatticeReturnsEmpty() throws Exception {
-            byte[] pdf = blankPdf();
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                assertTrue(parser.parse(doc, new RawPage(1, 0f, 0f, List.of())).isEmpty());
-            }
-        }
-    }
-
-    // ── stream mode (BasicExtractionAlgorithm) ───────────────────────────────
-
-    @Nested
-    @DisplayName("Stream mode")
-    class StreamMode {
-
-        @Test
-        @DisplayName("page with text yields at least one well-formed fragment")
-        void streamOnTextProducesFragment() throws Exception {
-            byte[] pdf =
-                    pdfWithText(new String[] {"Name Age City", "Alice 30 Paris", "Bob 25 Rome"});
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                List<TableFragment> fragments =
-                        parser.parseStream(doc, new RawPage(1, 0f, 0f, List.of()));
-                assertNotNull(fragments);
-                assertFalse(fragments.isEmpty(), "stream mode always builds a table from text");
-                assertFragmentWellFormed(fragments.get(0), 1, 0);
-            }
-        }
-
-        @Test
-        @DisplayName("fragment ids encode page and index")
-        void streamFragmentIdFormat() throws Exception {
-            byte[] pdf = pdfWithText(new String[] {"col1 col2", "a b"});
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                List<TableFragment> fragments =
-                        parser.parseStream(doc, new RawPage(1, 0f, 0f, List.of()));
-                assertFalse(fragments.isEmpty());
-                assertEquals("tbl-p1-0", fragments.get(0).tableId());
-                assertEquals(1, fragments.get(0).pageNumber());
-            }
-        }
-
-        @Test
-        @DisplayName("rawRows and the parsed rows stay in lockstep")
-        void streamRowsMatchRawRows() throws Exception {
-            byte[] pdf = pdfWithText(new String[] {"x y", "1 2", "3 4"});
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                List<TableFragment> fragments =
-                        parser.parseStream(doc, new RawPage(1, 0f, 0f, List.of()));
-                assertFalse(fragments.isEmpty());
-                TableFragment f = fragments.get(0);
-                assertEquals(f.rawRows().size(), f.rows().size());
-            }
-        }
-    }
-
-    // ── lattice mode with a real bordered grid ───────────────────────────────
-
-    @Nested
-    @DisplayName("Lattice mode")
-    class LatticeMode {
-
-        @Test
-        @DisplayName("bordered grid is detected and produces well-formed fragments")
-        void latticeDetectsBorderedTable() throws Exception {
-            byte[] pdf = pdfWithGrid();
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                List<TableFragment> fragments =
-                        parser.parse(doc, new RawPage(1, 0f, 0f, List.of()));
-                assertNotNull(fragments);
-                assertFalse(
-                        fragments.isEmpty(), "a clean ruled grid must be detected in lattice mode");
-                TableFragment f = fragments.get(0);
-                assertFragmentWellFormed(f, 1, 0);
-                assertTrue(f.columnCount() >= 1, "a detected grid must have at least one column");
-                assertFalse(f.rawRows().isEmpty(), "a detected grid must have rows");
-            }
-        }
-
-        @Test
-        @DisplayName("convenience overload with page number routes to lattice mode")
-        void parseByPageNumberDetectsGrid() throws Exception {
-            byte[] pdf = pdfWithGrid();
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                List<TableFragment> fragments = parser.parse(doc, 1);
-                assertNotNull(fragments);
-                assertFalse(fragments.isEmpty());
-                assertEquals(1, fragments.get(0).pageNumber());
-            }
-        }
-
-        @Test
-        @DisplayName("cell text is normalised (trimmed, newlines collapsed)")
-        void latticeCellTextIsNormalised() throws Exception {
-            byte[] pdf = pdfWithGrid();
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                List<TableFragment> fragments =
-                        parser.parse(doc, new RawPage(1, 0f, 0f, List.of()));
-                assertFalse(fragments.isEmpty());
-                for (List<String> row : fragments.get(0).rawRows()) {
-                    for (String cell : row) {
-                        assertNotNull(cell);
-                        assertFalse(cell.contains("\n"), "newlines must be collapsed");
-                        assertFalse(cell.contains("\r"), "carriage returns must be collapsed");
-                        assertEquals(cell.trim(), cell, "cell text must be trimmed");
-                    }
-                }
-            }
-        }
-    }
-
-    // ── contract invariants ──────────────────────────────────────────────────
-
-    @Nested
-    @DisplayName("Contract invariants")
-    class ContractInvariants {
-
-        @Test
-        @DisplayName("parse never returns null")
-        void parseNeverReturnsNull() throws Exception {
-            byte[] pdf = pdfWithText(new String[] {"abc"});
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                assertNotNull(parser.parse(doc, new RawPage(1, 0f, 0f, List.of())));
-                assertNotNull(parser.parse(doc, 1));
-                assertNotNull(parser.parseStream(doc, new RawPage(1, 0f, 0f, List.of())));
-            }
-        }
-
-        @Test
-        @DisplayName("the document is not closed by the parser")
-        void documentRemainsOpenAfterParse() throws Exception {
-            byte[] pdf = pdfWithText(new String[] {"keep me open"});
-            try (PDDocument doc = Loader.loadPDF(pdf)) {
-                parser.parse(doc, new RawPage(1, 0f, 0f, List.of()));
-                parser.parseStream(doc, new RawPage(1, 0f, 0f, List.of()));
-                // ObjectExtractor.close() would close the underlying COSDocument; the parser must
-                // not.
-                assertFalse(
-                        doc.getDocument().isClosed(),
-                        "parser must not close the caller's document");
-                assertEquals(1, doc.getNumberOfPages());
-            }
-        }
-    }
-
-    // ── helpers ──────────────────────────────────────────────────────────────
-
-    /** Asserts every field of a fragment satisfies the documented contract. */
-    private static void assertFragmentWellFormed(
-            TableFragment f, int expectedPage, int expectedIndex) {
-        assertNotNull(f);
-        assertEquals(expectedPage, f.pageNumber());
-        assertEquals("tbl-p" + expectedPage + "-" + expectedIndex, f.tableId());
-        assertNotNull(f.bounds());
-        assertNotNull(f.headers());
-        assertTrue(f.headers().isEmpty(), "headers are deferred to v2 and must be empty");
-        assertNotNull(f.rows());
-        assertNotNull(f.rawRows());
-        assertNotNull(f.warnings());
-        assertSame(null, f.continuedFromPage(), "continuedFromPage is deferred to v2");
-        assertTrue(f.columnCount() >= 0);
-        assertTrue(f.confidence() >= 0f && f.confidence() <= 1f, "confidence must be within [0,1]");
-        assertEquals(f.rawRows().size(), f.rows().size());
-
-        for (TableRow row : f.rows()) {
-            assertNotNull(row.cells());
-            for (TableCell cell : row.cells()) {
-                assertNotNull(cell.text());
-                assertNotNull(cell.bounds());
-                assertEquals(1, cell.colSpan(), "colSpan is always 1 in v1");
-                assertEquals(1, cell.rowSpan(), "rowSpan is always 1 in v1");
-            }
-        }
-    }
-
-    private static byte[] pdfWithText(String[] lines) throws Exception {
-        try (PDDocument doc = new PDDocument()) {
-            PDPage page = new PDPage(PDRectangle.A4);
-            doc.addPage(page);
-            try (PDPageContentStream cs = new PDPageContentStream(doc, page)) {
-                cs.setFont(new PDType1Font(Standard14Fonts.FontName.HELVETICA), 12);
-                cs.setNonStrokingColor(Color.BLACK);
-                float y = 720f;
-                for (String line : lines) {
-                    cs.beginText();
-                    cs.newLineAtOffset(72f, y);
-                    cs.showText(line);
-                    cs.endText();
-                    y -= 20f;
-                }
-            }
-            return save(doc);
-        }
-    }
-
-    private static byte[] blankPdf() throws Exception {
-        try (PDDocument doc = new PDDocument()) {
-            doc.addPage(new PDPage(PDRectangle.A4));
-            return save(doc);
-        }
-    }
-
-    /**
-     * Builds a small 3-row x 3-column ruled grid with text in each cell. The ruled lines make the
-     * table detectable by lattice mode.
-     */
-    private static byte[] pdfWithGrid() throws Exception {
-        try (PDDocument doc = new PDDocument()) {
-            PDPage page = new PDPage(PDRectangle.A4);
-            doc.addPage(page);
-
-            float left = 100f;
-            float right = 400f;
-            float top = 700f;
-            float bottom = 550f;
-            int cols = 3;
-            int rows = 3;
-            float colStep = (right - left) / cols;
-            float rowStep = (top - bottom) / rows;
-
-            try (PDPageContentStream cs = new PDPageContentStream(doc, page)) {
-                cs.setStrokingColor(Color.BLACK);
-                cs.setLineWidth(1f);
-
-                // vertical lines
-                for (int c = 0; c <= cols; c++) {
-                    float x = left + c * colStep;
-                    cs.moveTo(x, bottom);
-                    cs.lineTo(x, top);
-                }
-                // horizontal lines
-                for (int r = 0; r <= rows; r++) {
-                    float yLine = bottom + r * rowStep;
-                    cs.moveTo(left, yLine);
-                    cs.lineTo(right, yLine);
-                }
-                cs.stroke();
-
-                // cell text
-                cs.setFont(new PDType1Font(Standard14Fonts.FontName.HELVETICA), 10);
-                cs.setNonStrokingColor(Color.BLACK);
-                for (int r = 0; r < rows; r++) {
-                    for (int c = 0; c < cols; c++) {
-                        cs.beginText();
-                        cs.newLineAtOffset(left + c * colStep + 5f, top - (r + 1) * rowStep + 6f);
-                        cs.showText("R" + r + "C" + c);
-                        cs.endText();
-                    }
-                }
-            }
-            return save(doc);
-        }
-    }
-
-    private static byte[] save(PDDocument doc) throws Exception {
-        ByteArrayOutputStream baos = new ByteArrayOutputStream();
-        doc.save(baos);
-        return baos.toByteArray();
-    }
-}
--- a/Show More
+++ b/Show More