ai-prlisted
Install: claude install-skill arcasilesgroup/ai-engineering
# PR Workflow
Governed PR creation: full commit pipeline, pre-push gates, structured PR with summary + test plan, auto-complete with squash merge and branch deletion.
`/ai-pr` is the canonical chain step `brainstorm → plan → build → pr` — it runs the commit pipeline internally; the operator does NOT run `/ai-commit` first. `/ai-commit` exists as a standalone off-chain skill for WIP-only commits where no PR is wanted.
```
/ai-pr # full pipeline + create or update PR
/ai-pr --draft # open as draft (no review request)
/ai-pr review # request review on existing PR
/ai-pr update # refresh PR body + push amended commit
```
## Process
### Steps 0-6: Inline Commit Pipeline
Step 0 — **Auto-branch** from `main`/`master`: infer type (`feat/`, `fix/`, `chore/`, `docs/`, `refactor/`), generate slug with `python3 .ai-engineering/scripts/branch_slug.py --prefix <type>`, then `git checkout -b <output>`.
Step 1 — **Work-item context (opt)**: spec.md frontmatter `refs` → commit body trailers (`Refs: AB#101, #45`); only `close_on_pr` items, never features.
Step 2 — **Instinct consolidation**: if `.ai-engineering/observations/observations.yml` exists, run `/ai-session-watch --review` before committing.
Step 3 — **Stage** selectively (`git add <file>...`). Use `git add -A` only when explicitly requested. Exclude generated files, secrets, large binaries.
Step 4 — **Run gate orchestrator**: `ai-eng gate run --cache-aware --json --mode=local`. The