4 commits

Author	SHA1	Message	Date
lakshit verma	35069fe4ad	fix(backend): parse triage proposals wrapped in prose or code fences ... Free-model routers often answer with surrounding prose, so a strict whole- response JSON parse failed and every triage run ended in error. Extract the first JSON block before parsing, keeping the zod schema gate intact.	2026-08-19 22:05:20 +05:30
lakshit verma	1ca37a6504	feat(backend): return the created run id from manual triage ... runTriage now resolves to { outcome, runId } so the admin API can hand the run id back to the UI for deep-linking into the run detail.	2026-08-19 21:46:21 +05:30
lakshit verma	b8cb205e69	test(backend): raise AI surface coverage — provider client, admin agent API, policy, triage skip paths	2026-08-19 20:03:39 +05:30
lakshit verma	7a84996f51	test(backend): agent runtime, copilot API, and triage/worker suites with mocked provider ... Also fix two bugs the tests surfaced in the merged AI implementation: - agent-tool-call model made approval.summary/expiresAt conditionally required only when an approval subdocument is present, so blocked/error tool calls persist without a fake approval - agent.repo.createRun coerced empty-string session/user ids to null so autonomous triage runs (no session/user) no longer fail ObjectId casting	2026-08-19 19:23:00 +05:30