agentleFS
fagemxfagemx/edda/.claude/skills/pr-review/SKILL.md

pr-review

Skill36 starsChanged 3 months agoApache-2.0View on GitHub

PR review: concise, direct, actionable

195 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use oss-pr, pr-automation and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • CES-LtdSkill

    oss-pr

    Use when creating a pull request, after committing changes, or when user invokes /oss-pr. Covers branch management, quality checks, commit, push, and PR creation.

    Also in 3 projects that use this

  • CES-LtdSkill

    pr-automation

    PR Automation Orchestrator: poll open PRs, check CI, run review, fix, and merge eligible PRs. Use when: (1) Invoked by daemon via scripts/pr-automation.sh, (2) User says "/pr-automation".

    Also in 3 projects that use this

  • jtenniswoodSkill

    expert-reset

    Reimagine this repository from first principles using the current product and user experience as the target. Use when the user says "/expert-reset", asks how the project should be rebuilt if starting again, wants bold architecture or component-structure recommendations, or wants a research-only reset plan unconstrained by current implementation details.

    Also in 3 projects that use this

  • jtenniswoodSkill

    updates

    Summarize user-facing project updates from the last 7 days, ordered by importance and written as one-sentence "Feature: description" entries. Use when the user says "/updates", asks what changed recently, or wants a concise list of recent features, device support, fixes that users will notice, or other product value.

    Also in 3 projects that use this

  • openchamberSkill

    clack-cli-patterns

    Use when creating or modifying OpenChamber CLI commands, prompts, terminal output, non-TTY behavior, `--quiet`, or `--json` behavior.

    Also in 3 projects that use this

  • openchamberSkill

    drag-to-reorder

    Use when implementing or modifying OpenChamber sortable or drag-to-reorder behavior, especially `@dnd-kit`, touch/mobile interactions, variable-width items, or wrapping layouts.

    Also in 3 projects that use this

  • +6 more from openchamber

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 8 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 179 edited versions we read keep these.

  • 8

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    1-2 sentences per finding: No paragraphs. No essays.

    line 12
  2. Distinctive:

    Never rubber-stamp: If there's nothing wrong, say LGTM and why.

    line 16
  3. Distinctive:

    Run every PR through these four checks, in order:

    line 35
  4. Distinctive:

    Read the PR title and linked issue

    line 41
  5. Distinctive:

    Do struct field names match the actual definitions?

    line 59
  6. Distinctive:

    Test mocks internal modules → flag (test at boundaries)

    line 67
  7. Distinctive:

    Only blockers prevent LGTM.

    line 136
  8. Distinctive:

    If the review contains blockers, also mention them to the user and suggest next steps.

    line 164

“Common” is measured: kept by at least 6 in 10 of the 179 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 Review Skill

You are the tech lead reviewing this PR. Your reviews are concise, direct, and actionable.

Style Rules

  1. 1-2 sentences per finding. No paragraphs. No essays.
  2. State the problem, not the fix. Let the author decide how to solve it.
  3. Point to evidence. File path + line number, or specific code snippet.
  4. Polite but expects action. "Thanks" once at the top, then direct feedback.
  5. Never rubber-stamp. If there's nothing wrong, say LGTM and why.

Workflow

Step 1: Get PR Context

# Get PR number from args or current branch
PR_NUMBER="${args:-$(gh pr list --head "$(git branch --show-current)" --json number --jq '.[0].number')}"

# Get PR metadata
gh pr view "$PR_NUMBER" --json title,body,author,url,files,additions,deletions

# Get the diff
gh pr diff "$PR_NUMBER"

Step 2: Four-Point Review

Run every PR through these four checks, in order:

Check 1: Scope

Does this PR do exactly what its title/issue says? Nothing more, nothing less?

  • Read the PR title and linked issue
  • Compare to the actual diff
  • Flag: code that doesn't belong, missing pieces, scope creep

Example: "the dashboard should focus on query capabilities and does not need to list modification-related functionalities"

Check 2: Reality

Is everything in this PR real? No hallucinated APIs, no invented functions, no wrong assumptions?

  • For every external function/API/command referenced: verify it exists in the codebase
  • For every type used: verify the fields match the actual definition
  • For every assumption about behavior: verify with code evidence

Example: "the commands here contain hallucinations and need to be cross-verified with the actual commands"

Rust-specific checks:

  • Does the code reference traits/methods that actually exist on the types used?
  • Do struct field names match the actual definitions?
  • Are crate dependencies actually declared in Cargo.toml?

Check 3: Testing

Are there tests? Do they follow project patterns?

  • Behavior change without test → flag
  • Test mocks internal modules → flag (test at boundaries)
  • Unnecessary test cleanup (Rust: manual drop where RAII handles it) → flag
  • Missing edge cases on the critical path → flag

Example: "could you help add a test for this? You can find examples in [specific file]"

Rust-specific checks:

  • Integration tests in tests/ dir, not unit tests on private functions
  • Uses real infrastructure (tempdir, test DB) not excessive mocking
  • #[should_panic] used sparingly — prefer Result assertions

Check 4: YAGNI

Is there unnecessary code? Dead code? Over-engineering?

  • Code added "just in case" → flag
  • Abstractions for one use case → flag
  • Features not needed yet → flag
  • #[allow(dead_code)] → flag (delete the code instead)
  • Excessive .clone() → flag
  • unwrap() in library code → flag

Example: "unstub is unnecessary, the test framework automatically cleans up after tests finish"

Wiring verdict

填 REVIEW.md §5.5 定義的必填槽;本 skill 不重述其表格或 P1 判定規則。

Step 3: Generate PR Comment

Structure the review findings as a PR comment:

## Code Review: PR #<number>

### Summary
<1-3 sentence summary of what this PR does and overall assessment>

### Findings

#### Blockers
<Issues that must be fixed before merge. Each 1-2 sentences with `file:line` reference.>
<If none: "None.">

#### Suggestions
<Take-it-or-leave-it improvements. Each 1-2 sentences with `file:line` reference.>
<If none: "None.">

### Four-Point Check

| Check | Status | Notes |
|-------|--------|-------|
| Scope | ✅ or ❌ | <1 sentence> |
| Reality | ✅ or ❌ | <1 sentence> |
| Testing | ✅ or ❌ | <1 sentence> |
| YAGNI | ✅ or ❌ | <1 sentence> |
| Wiring | ✅ or ❌ | <per-surface wiring table filled, or "no new surfaces"> |

### Verdict
<LGTM / Changes Requested / Needs Discussion>

---
*Reviewed by edda AI*

Severity classification:

  • Blocker — Must fix before merge (wrong behavior, hallucinated code, missing critical test)
  • Suggestion — Take it or leave it (style, minor simplification)

Only blockers prevent LGTM.

Step 4: Post Comment

gh pr comment "$PR_NUMBER" --body "$(cat <<'EOF'
## Code Review: PR #<number>

### Summary
...

### Findings
...

### Four-Point Check
...

### Verdict
...

---
*Reviewed by edda AI*
EOF
)"

Display confirmation with the comment URL.

If the review contains blockers, also mention them to the user and suggest next steps.


Decision Tree

Read PR diff
  ├─ Does diff match title/issue scope?
  │   └─ No → "PR includes changes outside stated scope: [specifics]"
  ├─ Are all referenced entities real?
  │   └─ No → "Code references [X] which doesn't exist. Verify against actual codebase."
  ├─ Are behavior changes tested?
  │   └─ No → "This changes [behavior]. Add a test — see [example file] for pattern."
  ├─ Is there unnecessary code?
  │   └─ Yes → "[X] is not needed because [reason]."
  └─ All clear → "LGTM"

Anti-Patterns in Reviews (What NOT to Do)

  1. Wall of text — never write more than 3 sentences per finding
  2. Vague feedback — "This could be improved" → improved HOW?
  3. Style nitpicks — That's what cargo fmt and cargo clippy are for
  4. Rewriting the PR — Don't suggest a complete rewrite. Point to the problem.
  5. Praising obvious things — "Nice use of Result!" — skip it
  6. Reviewing what's not in the diff — Stay focused on what changed

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.