ai-session-watch-sweep

Solid

Consolidates the session-watch observation backlog on demand: runs /ai-session-watch --review, gates the result, and opens a draft chore PR for human review. Trigger for 'run the observation sweep', 'consolidate observations', 'session-watch sweep'. Never auto-merges, never auto-files work items, never runs unattended. Not for interactive review; use /ai-session-watch --review instead.

AI & Automation 52 stars 3 forks Updated 5 days ago MIT

Install

View on GitHub

Quality Score: 89/100

Stars 20%
57
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
100
Issue Health 10%
80
License 10%
100
Description 5%
100

Skill Content

# Session-Watch Sweep ## Purpose The session-watch `corrections` loop only consolidates when a human runs `/ai-session-watch --review` — it has no reliable trigger, so operator corrections accumulate unconsolidated (spec-165). This skill is a manual wrapper that runs the review, gates the result, and opens a **draft chore PR** so a human reviews the consolidated lessons before merge — keeping consolidation off feature branches. It has no scheduler: an operator runs it, and the SessionStart observation-nudge (spec-165 D-165-03) surfaces the pending backlog as a reminder. ## When to Use - Periodic manual consolidation pass (weekly cadence recommended). - Before a release-cut to capture accumulated session lessons. - NOT for interactive review — that should call `/ai-session-watch --review` directly (which keeps work-item creation). ## Hard Rules - Never auto-merge. The PR is always opened with `--draft`. - Never auto-file work items. The sweep runs `--review`'s extract → enrich → write steps only; the human decides what becomes an issue from the PR (spec-165 D-165-04). Unattended issue filing is forbidden. - If the review produces no corpus change, exit cleanly with a status event; do NOT open an empty PR. ## Process ### Step 1 — Run the review (work-item creation suppressed) Invoke the consolidation in non-interactive mode: ``` /ai-session-watch --review ``` Run only steps 1-3 (extract → enrich → write) plus the `lastReviewedAt` stamp; SKIP step 5 (create wo...

Details

Author
arcasilesgroup
Repository
arcasilesgroup/ai-engineering
Created
4 months ago
Last Updated
5 days ago
Language
Python
License
MIT

Similar Skills

Semantically similar based on skill content — not just same category