M1 — TOTP secrets are now AES-256-GCM encrypted before being stored in
the database. Key is auto-generated on first run (data/totp.key) or set
via OWNCORD_TOTP_KEY env var. Existing plaintext secrets are detected
and returned as-is for backwards compatibility.
M3 — Replay buffer events are now tagged with their channel ID. On
reconnect, the server computes the user's current accessible channels
and only replays events from those channels. Global broadcasts (presence,
voice state, member updates) are always replayed. Falls back to full
ready payload if permission computation fails.
Security audit across all 11 sections (AUTH-001 through DATA-001) found
0 critical, 1 high, 7 medium, 15 low issues. This commit addresses:
- Add json:"-" to User.PasswordHash, User.TOTPSecret, Session.TokenHash
to prevent accidental serialization of sensitive fields (M7)
- Add X-Content-Type-Options: nosniff to file serve responses (M5)
- Apply owner-only guard to backup list endpoint for consistency (M6)
- Persist rate-limit lockouts to SQLite so they survive restarts (M2)
- Normalize DM non-participant responses to 404 to prevent oracle (L3)
- Add explicit per-entry expiry check in partial auth Lookup/Consume (L1)
- Truncate unknown WS message type to 64 chars before echo (L6)
- Rate-limit ping handler to 2/sec per user (L7)
- Replace raw error strings in update handlers with generic messages (L15)
- Update 4 tests to match new 404 behavior for DM non-participant
Changed IsCA to false, removed KeyUsageCertSign, and reduced validity
from 10 years to 2 years. A compromised key can no longer sign
additional certificates trusted by TOFU-pinning clients.
- Add syncutil package with deadlock-detecting mutex (build-tag switchable)
- Add main_test.go TestMain scaffolding across all server packages
- Harden concurrency in ws, admin, auth, and updater packages
- Update CI workflow, go.mod/sum, Cargo.lock, and root changelogen tooling
Create Server/api/constants.go (26 constants) and Server/auth/constants.go
(6 constants) for rate limits, timeouts, size limits, and token generation.
Replace all inline magic numbers with descriptive names across 9 source files.
No behavior changes — same values, just named for contributor readability.
CRITICAL (5):
- Hub panic recovery now calls h.Stop() after 3 panics (ws/hub.go)
- Ring buffer EventsSince returns non-nil empty slice for current seq (ws/ringbuffer.go)
- PTT event listener stores unsubscribe handle to prevent leak (ptt.ts)
- verifyTotp respects config.allowSelfSigned instead of hardcoding (api.ts)
- ptt_listen_for_key uses spawn_blocking to avoid thread pool starvation (ptt.rs)
HIGH - Server (13):
- TOTP rate-limit checked after body decode; counters reset on success
- TOTP enable returns 409 if already enabled (must disable first)
- Global search pre-computes accessible channel IDs for FTS WHERE clause
- DeleteAccount queries roles by name instead of hard-coded IDs
- BackupToSafe uses absClean in VACUUM INTO
- Voice camera slot uses atomic EnableCameraIfUnderLimit DB method
- readPump snapshots voiceChID before unregister for TOCTOU safety
- Voice join sets state after token send; rollback takes broadcast flag
- Updater download uses probe pattern instead of overflow write
- Webhook checks Authorization header before reading body
- Storage.Save adds fsync and fixes double-close
- Default WS origin denies cross-origin (was: accept all)
HIGH - Client (6):
- WS reconnect uses generation counter to discard stale events
- AudioPipeline uses generation counter against stale worklet callbacks
- Screenshare mute state preserved across reconnect (not full leave)
- handleVoiceToken uses iterative loop instead of unbounded recursion
- store.ts re-entrancy guard with pending update queue
- Notification AudioContext cleaned up on logout
Reviewed by 4 parallel agents across Server Core, Server Realtime,
Client & Tauri, and Security. 55 total findings; 24 CRITICAL+HIGH
fixed here, 31 MEDIUM+LOW tracked in vault backlog (T-265–T-295).
- C-3: inject setupLimiter into NewAdminAPI instead of package-level global
- H-5: generateRandomKey returns error instead of panicking
- H-6: replace init() bcrypt with sync.Once lazy initialization
- M-2: remove unsafe-inline from admin CSP
- M-3: sanitize upload filenames (strip control chars, truncate to 255)
- M-5: truncate User-Agent to 512 bytes before storing as device
- M-10: MaxBodySizeUnless uses prefix matching instead of exact path
- M-12: wrap seedExistingDatabase in a single transaction
- M-13: use errors.Is for EOF check in upload handler
- M-14: log writeJSON encoding errors instead of discarding
- M-16: standardize error codes to INTERNAL_ERROR across all handlers
- C-3: inject setupLimiter into NewAdminAPI instead of package-level global
- H-5: generateRandomKey returns error instead of panicking
- H-6: replace init() bcrypt with sync.Once lazy initialization
- M-2: remove unsafe-inline from admin CSP script-src and style-src
- M-3: sanitize upload filenames (strip control chars, truncate to 255)
- M-5: truncate User-Agent to 512 bytes before storing as device
- M-10: MaxBodySizeUnless uses prefix matching instead of exact path
- M-12: wrap seedExistingDatabase in a single transaction
- M-13: use errors.Is for EOF check in upload handler
- M-14: log writeJSON encoding errors instead of discarding
- M-16: standardize error codes to INTERNAL_ERROR across all handlers
Server fixes:
- voice_leave: broadcast voice_leave even on DB error so peers don't see ghost users (H1)
- migrate: record migration inside transaction for atomicity (H2)
- password: use init() with panic for dummyHash to catch bcrypt init failures (M1)
- password: replace //nolint:errcheck with explicit _ = discard (M2)
- handlers: clarify edit permission comment, fix error message wording (M3)
- auth_handler: distinguish duplicate username (400) from DB error (500) (M4)
Client fixes:
- media: revert YouTube oEmbed to browser fetch — no need to disable cert verification (C1)
- messages.store: fix prependMessages cap to keep newest messages, not oldest (H5)
- ChannelSidebar: ref-count globalDragAc to prevent multi-instance teardown race (H6)
- attachments: replace console.error with project logger (M6)
- ws: clarify lastSeq reset comment to match actual behavior (M5)
Fix 70+ errcheck violations by adding explicit error discards (`_ =`)
for unchecked return values across test helpers and deferred Close()
calls. Remove unused `senderID` field from broadcastMsg and unused
`defaultCleanupMaxWindow` const. Apply De Morgan's law, remove empty
branch, and simplify redundant type declaration per staticcheck.
- Scaffold Go module (github.com/owncord/server) with all package dirs
- config: koanf-based YAML loader with env var overrides, default generation
- db: pure-Go SQLite (modernc, no CGO), WAL mode, FK enforcement, full
15-table schema from SCHEMA.md including FTS5 and idempotent migrations
- auth/tls: ECDSA P-256 self-signed cert generation, LoadOrGenerate for
all 4 TLS modes (self_signed, acme, manual, off)
- api: chi router with request ID middleware, /health and /api/v1/info
- main: graceful shutdown (30s timeout), structured slog JSON logging
- Stubs for ws, storage, admin packages ready for Phase 2+
Test coverage: api 100%, auth 85.7%, db 82.4%, config 80.6%
Binary: chatserver.exe 12MB, GOOS=windows GOARCH=amd64
J3vb