No description
Forgejo signal source tests: - Signal parsing for empty body, mixed content, large issue numbers - findLinkedPR edge cases (multiple refs, empty body, first match wins) - Poll with combined status failure/error fallback to PENDING - Poll with unassigned child (no signal), child fetch failure (skipped) - Poll with multiple epics, mixed labels (only epics generate signals) - Report format verification (success/failure, error in code block) Journal replay tests: - Write + read-back round-trip with full field preservation - Filter by action (tick_parent from mixed actions) - Filter by repo (date-partitioned files in owner/repo directories) - Filter by time range (entries across multiple days) - Combined filter (action + repo + time) - Concurrent write safety (20 goroutines, valid JSONL output) - Empty journal returns no entries Handler integration tests: - TickParent full flow: signal -> execute -> verify epic body updated, child closed, progress tracked (checked/unchecked counts) - EnableAutoMerge: verify squash merge method via mock - PublishDraft: verify draft=false PATCH sent - SendFixCommand: conflict vs review message paths - Completion: success flow (label swap, comment) and failure flow - Full pipeline: signal -> handler -> journal write -> read-back - Handler priority: verify correct handler matches each signal state (merged, draft, conflicting, open+passing) Co-Authored-By: Virgil <virgil@lethean.io> |
||
|---|---|---|
| agentci | ||
| collect | ||
| forge | ||
| git | ||
| gitea | ||
| jobrunner | ||
| CLAUDE.md | ||
| FINDINGS.md | ||
| go.mod | ||
| go.sum | ||
| TODO.md | ||