agentleFS
aiconnaiaiconnai/engram/skills/pr-review-triage/SKILL.md

pr-review-triage

Skill3 starsChanged 10 months agoMITView on GitHub

Use for an Engram PR babysitter triage loop that monitors PR aging, CI blocks, reviewer threads, and lightweight action proposals with isolated-worktree fix candidates and human escalation for risky changes.

3 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 6 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 6

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  • could be automatic: a setting could enforce it
  1. Very specific:

    Use this skill for PR queue hygiene on a weekly or 5 to 15 minute cadence.

    line 8
  2. Distinctive:

    Only propose an isolated-worktree fix when all of these are true:

    line 44
  3. Distinctive:

    No actionable PRs: exit report-only after updating state and run log.

    line 57
  4. Distinctive:

    Do not auto-merge without an explicit allowlist.

    line 61
  5. Distinctive:

    Do not resolve reviewer comments unless the verifier confirms the diff and intent.

    line 62
  6. Distinctive:

    Do not edit auth, storage migrations, protocol surfaces, attestation/snapshot code, production infrastructure, secrets, release workflows, …

    line 64
  7. Distinctive:

    Prefer a short comment or issue handoff over speculative code changes.

    line 67

“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 Babysitter Triage Loop

Use this skill for PR queue hygiene on a weekly or 5 to 15 minute cadence. The default mode is proposal-only unless a low-risk L2 fix candidate passes the gate. Pair it with skills/loop-engineering/SKILL.md for global loop controls.

Inputs

  • Open PR list and labels.
  • CI status per PR.
  • Required reviewer threads and blocking comments.
  • Current pr-babysitter-state.md, history, run log, and budget file.
  • Current denylist and auto-merge allowlist.

Output

Append a concise triage block:

## High Priority
- PR:
  Blocking condition:
  Why it blocks delivery:
  Exact action:
  Risk level:

## Watch

## Human Escalation

## State Updates
- Attempts already made:
- Last action taken:
- Who or what is next:

Execution Gate

Only propose an isolated-worktree fix when all of these are true:

  • CI is red or there is a clear review block.
  • Attempts are below 3.
  • The issue is a single-file or small-diff candidate.
  • The path is not denylisted.
  • A verifier is available after the implementer.

Decision Matrix

  • Low-risk minimal fix candidate: spawn minimal-fix, then loop-verifier.
  • Medium or high risk, ambiguity, or path risk: append to triage inbox and escalate to a human.
  • No actionable PRs: exit report-only after updating state and run log.

Rules

  • Do not auto-merge without an explicit allowlist.
  • Do not resolve reviewer comments unless the verifier confirms the diff and intent.
  • Do not edit auth, storage migrations, protocol surfaces, attestation/snapshot code, production infrastructure, secrets, release workflows, or MCP write tools without human approval.
  • Prefer a short comment or issue handoff over speculative code changes.

Completion Gate

Mark the run complete only when every high-priority PR has a next action, risk level, owner or escalation path, and run-log entry.

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.