Repository navigation
User-frontend defect sweep (11): missing routes, auth-timeout logout, Company OS gating, i18n, stale feed, indexing #10502
Description
Batch of 11 user-frontend defects (triaged live 2026-06-26).
Backend routes / health / config
- Docs improvements #1
/api/healthHEAD probe — CacheBuster marks backend offline (HEAD /api/health is 200 live; verify the actual CacheBuster URL/path) - Fix frontend errors and remove unused code #3
/api/agents/status404 — Agent Activity Monitor; router unregistered or wrong path (shows 3/5 agents) - Dev new gui #4
/api/dev-speedup/templates+/history404 —api/development_speedup.pyexists but not wired; register or graceful empty-state - Add comprehensive codebase analysis reports #11 Audit dashboard backend address shows
0.0.0.0(bind addr, not real host) — return/display real host
Frontend UX / resilience
- Docs improvements #2
/api/auth/me>5s → forced logout + WS token loss; raise timeout / add grace+retry before redirect - Dev new gui #5 Company OS 503 in single_user — gate nav + graceful 'requires company mode' message (NOT enable)
- Analysis report 20250803 #8 Knowledge Base shows raw i18n keys (Title/Subtitle/Search placeholder…) — fix key resolution
- Dev new gui #7 Live Activity Feed stale entry (1357h ago) + no refresh interval (endpoint returns 200)
- Add comprehensive codebase analysis reports #10 Security Analytics 'No Services' — config a provider or replace with setup guide
Runtime / data (investigate)
- Refactor backend API and remove unused variables #6 Error Handler agent stuck 'Error' status, 78% success, active 'Task failed' events
- Analysis report 20250814 1 #9 AutoBot-AI codebase 'Never indexed' (branch Dev_new_gui) — trigger initial index / investigate
Activity
fix(api): wire agents/status + dev-speedup routes, audit host, health…
fix(ui): Company OS gating, live-feed refresh, knowledge i18n, securi…
fix(llc/frontend): never surface raw HTTP 503 to end users (harden #1…
deploy(backend): single_company mode — surfaces Company OS + enables …
fix(types): clear vue-tsc-baseline — drop dead api-contract search al…
mrveiss commentedon Jul 2, 2026
Live re-triage vs current codebase (2026-07-02)
Re-verified all 11 items with file:line evidence. 8 already fixed since the 2026-06-26 triage (routing renames + #10794 + earlier #10502 follow-up commits), 3 are runtime/data (not code), and 1 real code fix — now in PR #10854.
Backend routes / health / config
useNetworkStatus.ts:72-78probesHEAD ${getApiBase()}/health; backend registers@app.head("/api/health")(endpoints.py:152); treats <500 as online, offline only after 2 fails./api/agents/status404 — ALREADY FIXED. Routeapi/agent_org.py:108, router mounted/agents(core_routers.py:486) →/api/agents/statusresolves; in api.ts. "3/5" was runtime data./api/dev-speedup/templates+/history404 — ALREADY FIXED. Routesdevelopment_speedup.py:625/643, mounted/dev-speedup(feature_routers.py:361)./historyempty on fresh system.0.0.0.0— STILL REAL → PR fix(audit): display routable host instead of bind-all 0.0.0.0 in audit dashboard (#10502 #11) #10854.AuditStatistics.vue:77rendersvm_source=config.vm.main= bind-all in prod. Fixed viaaudit_logger.display_host(resolves FQDN for bind-all/loopback).Frontend UX / resilience
companyOsEnabledfail-closed (App.vue:1012); localized 'unavailable' message; 503→unavailableflag (commits f80c28d, f1f3cdf).knowledge.search.*keys present en.json:5220+; parity test enforces other locales.collaboration/activity_broadcastpush + 30s heartbeat (useSessionCollaboration.ts:541,363). Stale entry = runtime WS-connectivity symptom.ThreatIntelligenceDashboard.vue:17-34already shows a setup guide + CTA for the no-provider state.Runtime / data (not code)
getSampleAgents()fallback preview (useAgentActivityData.ts:140); real data via /api/agents/status. No code defect.Plan: once #10854 merges, the entire code-actionable scope of this umbrella is complete. Items #6/#9/#10 are operational states (run the indexer / configure a provider / real agent data), not code bugs — recommend closing this issue on #10854 merge and, if desired, tracking #6/#9/#10 as ops notes rather than code defects.
fix(audit): show routable host not bind-all 0.0.0.0 in audit dashboard (
mrveiss commentedon Jul 2, 2026
Closed — all code-actionable items resolved. Item #11 (audit dashboard bind-address) merged to Dev_new_gui in #10854 (
180ed8594). Per the live re-triage above: #1,#2,#3,#4,#5,#7,#8 were already fixed since the 2026-06-26 triage (verified with file:line), and #6/#9/#10 are runtime/data states (sample-agent fallback / index-not-yet-run / no-provider-configured — each already has a graceful UI), not code defects. Nothing code-level remains.