← ClaudeAtlas

pr-stacklisted

Plan and execute splitting a large PR or branch into a reviewable stacked PR sequence.
rp1-run/rp1 · ★ 38 · AI & Automation · score 77
Install: claude install-skill rp1-run/rp1
# PR Stack Plan and execute splitting a large PR, local branch, or remote branch into an ordered stack of smaller reviewable PRs. Use the pre-resolved `projectRoot`, `kbRoot`, `workRoot`, and `codeRoot` values from the generated Workflow Bootstrap section. Run all source-control commands from `codeRoot`, and write workflow artifacts only under `workRoot`. Constraint: plan first; create branches/PRs only after explicit user approval. ## STATE-MACHINE ```mermaid stateDiagram-v2 [*] --> planning planning --> executing : approved planning --> reporting : plan_only executing --> reporting : stack_created reporting --> [*] : done ``` **On each phase transition**, report via: ```bash rp1 agent-tools emit \ --workflow pr-stack \ --type status_change \ --run-id {RUN_ID} \ --step {CURRENT_STATE} \ --data '{"status": "running"}' ``` - `RUN_ID` comes from the generated Workflow Bootstrap section. - Derive `RUN_NAME` from the resolved source: use `"PR Stack: PR #{number}"` for a GitHub PR, otherwise `"PR Stack: {source_ref}"`. - On the first emit only, include `--name "{RUN_NAME}"`. - On blocking error, emit the current step with `{"status":"failed","reason":"..."}` and stop. **Example sequence**: ```bash rp1 agent-tools emit --workflow pr-stack --type status_change --run-id {RUN_ID} --name "PR Stack: PR #42" --step planning --data '{"status": "running"}' rp1 agent-tools emit --workflow pr-stack --type waiting_for_user --run-id {RUN_ID} --step plann