agentleFS
akshaypimprikarakshaypimprikar/pragma/.claude/skills/pr-followup/SKILL.md

pr-followup

Skill4 starsChanged 6 days agoMITView on GitHub

Run the review chain for the PR's lane immediately after a PR is opened: /review alone for a pipeline PR, the built-in code-review then /review for an app PR, nothing for docs, release and sync PRs. Invoke right after a PR is created, or manually against an existing PR.

  • Commits and pushes

3 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 0 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 0

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use the built-in code-review skill, not the code-review:code-review plugin.

    line 17
  2. Very specific:

    Small pipeline PRs: when the lane is pipeline and the PR changes fewer than 30 lines (additions plus deletions, small_pr in …

    line 41
  3. Very specific:

    Only the app lane (which also covers a PR touching app and pipeline paths) runs both, because on pipeline diffs the second reviewer doubled …

    line 48
  4. Very specific:

    Run /review <PR>. It posts the verdict and updates the PR body's Review: line.

    line 59

“Common” is measured: kept by at least 6 in 10 of the 3 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

PR Followup Agent

Runs the review chain for the PR's lane immediately after a PR is opened, and records each result where the review-evidence CI check reads it (scripts/check_review_evidence.py). A pipeline PR gets /review alone. An app PR gets the built-in code-review (at medium effort) and then /review. The docs, release and sync lanes get neither.

On an app PR, code-review runs first so its fixes land before the /review rounds; a fix after an APPROVED verdict would otherwise need /review --confirm.

Use the built-in code-review skill, not the code-review:code-review plugin. The plugin starts 10-20 subagents per run (Haiku triage, 5 Sonnet reviewers, a Haiku scorer per issue); from 2026-09-28, when it became the default here, it made each PR's review several times more expensive. /review's isolated subagent already covers design and rule compliance. Note: code-review is a Claude Code skill and may not be invocable at all — either because it is unavailable in this session, or because this coding agent has no such skill mechanism. Either way, step 3 below catches the invocation error and continues to reporting rather than halting the whole command; you'll still need to run an equivalent review yourself before merging.

Trigger

Invoked right after gh pr create succeeds, or manually against an existing PR: /pr-followup 71 or /pr-followup fix/some-branch.

Process

  1. Run python3 scripts/check_pr_lane.py --git origin/<base> --head-branch <head> --base-branch <base>. The script prints the lane: app, pipeline, docs, release or sync. For the docs, release and sync lanes, report the lane and stop: they need neither step below. For the pipeline lane, or any PR that also touches pipeline paths (an app-lane PR still owes the pipeline lane's motivating_incident, per evidence_for_change in scripts/check_pr_lane.py), make sure the PR body has a non-empty Motivating incident: <what went wrong, with a link or date> line (or none (<reason>)); add it if missing, since review-evidence fails without it. Small pipeline PRs: when the lane is pipeline and the PR changes fewer than 30 lines (additions plus deletions, small_pr in scripts/pipeline_lanes.json) and touches none of its exclude_paths (the files that enforce the rules: hooks, gates, review, scripts/check_*, review_verdict.py, the install scripts, lane configs, workflows), skip /review too: the gates, the lane check and review-evidence decide, and review_verdict is not required. Report "small pipeline PR, no model review" and stop. Count with git diff --shortstat origin/<base>...HEAD. Lane routing: a pipeline PR gets one reviewer, so skip steps 2–3 and go to step 4. Only the app lane (which also covers a PR touching app and pipeline paths) runs both, because on pipeline diffs the second reviewer doubled the token cost and added design suggestions rather than defects (2026-10-05: the same 4-line Gate 0 fix got 7 inline comments across a project PR and pragma #100).
  2. app lane only: run the built-in code-review once: code-review <PR> medium --comment. Fix any issue it posts, commit and push. Then add or replace one line in the PR body with the head SHA it reviewed: code-review: <comment URL> at <sha>, or code-review: no issues at <sha> when it posted nothing.
  3. app lane only: if the step 2 invocation errors (e.g. Unknown skill), don't stall. Print ⚠️ code-review couldn't be invoked here (unavailable, or unsupported on this agent) — run an equivalent review yourself before merging. and continue.
  4. Run /review <PR>. It posts the verdict and updates the PR body's Review: line. Stop at CHANGES REQUESTED until the blocking issues are fixed. /review runs one round by default and a second only to confirm blocking fixes.
  5. Report the results (one for a pipeline PR, two for app) and the lane.

/test is not in this chain: it now runs between /feature and /gates.

Done when

Each result the lane needs is reported and recorded in the PR body (or the fallback warning printed). Do not merge — per AGENTS.md/CLAUDE.md's Merge rule, the required checks decide mergeability and the user merges.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.