Skip to content
Agentmart
Submit a tool

review-pr

Review a GitHub PR using the your-pr-reviewer agent with wiki-backed institutional knowledge.

Skill by agent-rails

IndexedApache-2.0
Write a review

Chalk Labs

Need help deploying review-pr?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/agent-rails-review-pr

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Review a GitHub PR using the your-pr-reviewer agent with wiki-backed institutional knowledge. Use when user provides a PR URL or says "review pr", "review this pr", "code review".

Use cases: agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/review-pr

What's inside

Skill name
review-pr
Description
Review a GitHub PR using the your-pr-reviewer agent with wiki-backed institutional knowledge. Use when user provides a PR URL or says "review pr", "review this pr", "code review".
License
Apache-2.0
Instructions
66 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

Review the PR using the `your-pr-reviewer` agent ecosystem.

## Input

The user provides one of:
- Full URL: `https://github.com/your-org/some-repo/pull/42`
- Shorthand: `your-org/some-repo#42`
- Just a number (assumes current repo): `#42`

Parse the PR reference and pass it to your-pr-reviewer.

## Execution

1. **Evaluate cross-vendor cascade policy + prefetch PR data** (orchestrator-side, before spawning anyone):
   - Fetch in a single parallel Bash batch — combines cascade decision with downstream-agent prefetch so each agent does not re-run the same `gh` calls:
     ```bash
     gh pr view {n} --json title,body,author,files,additions,deletions,changedFiles,baseRefName,headRefName,headRefOid,state,labels
     gh pr diff {n} --repo {owner}/{repo}
     ```
   - Apply the skip rule from `~/your-pr-reviewer/wiki/conventions/cascade-default-policy.md`: your-cross-vendor-reviewer cascades unless ALL three hold — diff is docs-only, total LOC < 100, no file touches `.github/workflows|actions/**`, `actions/**`, `**/Dockerfile`, `**/terraform|helm|k8s|argocd|applicationsets/**`, or any always-cascade path.
   - If skip rule applies → spawn your-pr-reviewer only.
   - If skip rule does NOT apply → spawn your-pr-reviewer + your-cross-vendor-reviewer as peers in parallel; once both return, pass your-cross-vendor-reviewer's `YOUR_CROSS_VENDOR_REVIEWER_REVIEW:` block to your-pr-reviewer as `YOUR_CROSS_VENDOR_REVIEWER_INPUT:` for synthesis.
   - Surface the cascade decision in the user-facing review header (e.g. "cycle 1 — your-pr-reviewer only (docs-only diff, skip rule applied)" or "cycle 1 — your-pr-reviewer + cross-vendor").
   - **Pass the fetched data to each spawned agent as a `PR_PREFETCH:` block** in the prompt. Saves 5–15s per agent on cold start by avoiding redundant `gh` round-trips.

2. Spawn the `your-pr-reviewer` agent with this prompt:

```
Review this PR: {pr_reference}

PR_PREFETCH:
  pr_view: <gh pr view JSON inline>
  diff: |
    <gh pr diff output inline>

Steps:
1. Use the PR_PREFETCH block above — do not re-fetch unless PR_PREFETCH is absent. Optionally pull comments via `gh api repos/{owner}/{repo}/pulls/{n}/comments` if needed.
2. Read wiki context from ~/your-pr-reviewer/ for this repo, author, and relevant conventions
3. Analyze the diff — only flag production-impacting issues
4. Output review findings
5. Delegate to your-pr-review-scribe to update the wiki with findings from this review
```

3. **Verify grounding before presenting** (orchestrator-side anti-hallucination check):
   - your-pr-reviewer's response MUST start with a `GROUNDING_OK:` block listing `head_sha` and `changed_files`.
   - Cross-check `head_sha` against `gh pr view {n} --json headRefOid`. If they don't match, halt and re-run.
   - Cross-check `changed_files` against `gh pr view {n} --json files`. If your-pr-reviewer's list is empty, fabricated, or includes paths not returned by `gh pr view`, halt and re-run.
   - For each finding's `File:` citation, confirm the path is in the verified `changed_files` list. Drop any finding that cites a path outside the list (or escalate back to your-pr-reviewer for a re-run if more than one slips through — that signals a grounding failure, not a single noisy finding).
   - If your-pr-reviewer returned `GROUNDING_FAILED:` or no checkpoint at all, do NOT present findings to the user. Re-spawn your-pr-reviewer with explicit grounding facts pasted into the prompt (head SHA + file list from `gh pr view`).
   - Reason: prior runs hallucinated findings against fictional files. This check is cheap and catches it deterministically before the user sees noise.

4. Present the synthesized review to the user, **led by a Merge Decision Pack** — the one-screen verdict + decision drivers at the top, full findings below. See `~/your-pr-reviewer/wiki/conventions/merge-decision-pack.md` for the format.

5. Ask: **Post as GitHub review?** If yes, use `gh pr review` to submit.

## Post-Review Actions

After presenting the review, offer:
- `[Post]` — Submit as GitHub PR review via `gh pr review {number} --repo {owner}/{repo}`
- `[Comment]` — Post individual inline comments via `gh api`
- `[Draft fixes]` — Invoke `/draft-pr-fixes {pr}` to spawn `your-patch-drafter` (codex-backed cross-vendor patch drafter). Produces unified-diff patches per finding for cherry-pick. Never auto-applies.
- `[Skip]` — Don't post, wiki still gets updated

Drafting fixes is HITL by design — patches surface for the user's pick, then Claude applies the chosen ones under the standard permission model. Use after the user has read the findings and wants candidate fixes without writing them by hand.

From agent-rails/agentic-fabric, Apache-2.0.

More from agent-rails/agentic-fabric

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Jul 27, 2026 (3 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
Apache-2.0
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from agent-rails

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get review-pr running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying review-pr?

CallGet help
Cite this page

review-pr skill. Agent Mart. https://agentmart.dev/skills/agent-rails-review-pr (last updated 2026-10-24)