agent-reviews
Skill78 starsChanged 9 months agoApache-2.0View on GitHub
Review and fix PR review bot findings on current PR, loop until resolved. Fetches unanswered bot comments, evaluates each finding, fixes real bugs, dismisses false positives, and replies to every comment with the outcome.
- Commits and pushes
In this repository9 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- code-reviewerSkill
- code-simplifierSkill
- comment-analyzerSkill
- fix-pr-commentsSkill
- moder-web-devSkill
- pr-test-analyzerSkill
- silent-failure-hunterSkill
- type-design-analyzerSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
22 rules. About 0 are common, 17 are distinctive, and 5 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈17
Distinctive · estimate
This author's own choices. Worth a look.
5
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 25
If no PR exists, notify the user and exit.
- Distinctive: line 33
If zero comments are returned, print "No unanswered bot comments found" and skip to Phase 2.
- Distinctive: line 41
Read the referenced code and determine:
- Distinctive: line 44
FALSE POSITIVE: - Not actually a bug (intentional behavior, bot misunderstanding)
- Distinctive: line 45
UNCERTAIN: - Not sure; ask the user
- Distinctive: line 72
If FALSE POSITIVE: Do NOT change the code. Track the comment ID and the reason it's not a real bug.
- Very specific: line 74
If UNCERTAIN: Use AskUserQuestion. If the user says skip, track it as skipped.
- Distinctive: line 76
Do NOT reply to comments yet.
- Distinctive: line 80
After evaluating and fixing ALL unanswered comments:
- Distinctive: line 82
Run your project's lint and type-check
- Very specific: line 123
Run scripts/agent-reviews.js --watch --bots-only as a background task.
- Very specific: line 125
6b.: Use TaskOutput to wait for the watcher to complete (blocks up to 12 minutes).
- Distinctive: line 127
6c.: Check the output:
- Very specific: line 130
Use --detail <id> to read each new comment's full detail
- Distinctive: line 135
Stop looping and move to the Summary Report.
- Distinctive: line 141
After both phases complete, provide a summary:
- Distinctive: line 167
Every finding gets a response: - No silent ignores
- Very specific: line 172
Use AskUserQuestion when uncertain about a finding
- Distinctive: line 173
Don't guess on architectural or business logic questions
- Distinctive: line 177
Verify findings before fixing - bots have false positives
- Distinctive: line 178
Keep fixes minimal and focused - don't refactor unrelated code
- Distinctive: line 179
Ensure type-check and lint pass before committing
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Automatically review, fix, and respond to findings from PR review bots on the current PR. Uses a deterministic two-phase workflow: first fix all existing issues, then poll once for new ones.
Path note: All scripts/agent-reviews.js references below are relative to this skill's directory (next to this SKILL.md file). Run them with node.
Phase 1: FETCH & FIX (synchronous)
Step 1: Identify Current PR
gh pr view --json number,url,headRefNameIf no PR exists, notify the user and exit.
Step 2: Fetch All Bot Comments (Expanded)
Run scripts/agent-reviews.js --bots-only --unanswered --expanded
This shows only unanswered bot comments with full detail: complete comment body (no truncation), diff hunk (code context), and all replies. Each comment shows its ID in brackets (e.g., [12345678]).
If zero comments are returned, print "No unanswered bot comments found" and skip to Phase 2.
Step 3: Process Each Unanswered Comment
For each comment from the expanded output:
A. Evaluate the Finding
Read the referenced code and determine:
- TRUE POSITIVE - A real bug that needs fixing
- FALSE POSITIVE - Not actually a bug (intentional behavior, bot misunderstanding)
- UNCERTAIN - Not sure; ask the user
Likely TRUE POSITIVE:
- Code obviously violates stated behavior
- Missing null checks on potentially undefined values
- Type mismatches or incorrect function signatures
- Logic errors in conditionals
- Missing error handling for documented failure cases
Likely FALSE POSITIVE:
- Bot doesn't understand the framework/library patterns
- Code is intentionally structured that way (with comments explaining why)
- Bot is flagging style preferences, not bugs
- The "bug" is actually a feature or intentional behavior
- Bot misread the code flow
When UNCERTAIN — use AskUserQuestion:
- The fix would require architectural changes
- You're genuinely unsure if the behavior is intentional
- The "bug" relates to business logic you don't fully understand
- Multiple valid interpretations exist
- The fix could have unintended side effects
B. Act on Evaluation
If TRUE POSITIVE: Fix the code. Track the comment ID and a brief description of the fix.
If FALSE POSITIVE: Do NOT change the code. Track the comment ID and the reason it's not a real bug.
If UNCERTAIN: Use AskUserQuestion. If the user says skip, track it as skipped.
Do NOT reply to comments yet. Replies happen after the commit (Step 5).
Step 4: Commit and Push
After evaluating and fixing ALL unanswered comments:
- Run your project's lint and type-check
- Stage, commit, and push:
git add -A git commit -m "fix: address PR review bot findings {List of bugs fixed, grouped by bot}" git push - Capture the commit hash from the output.
Step 5: Reply to All Comments
Now that the commit hash exists, reply to every processed comment:
For each TRUE POSITIVE:
Run scripts/agent-reviews.js --reply <comment_id> "Fixed in {hash}. {Brief description of the fix}"
For each FALSE POSITIVE:
Run scripts/agent-reviews.js --reply <comment_id> "Won't fix: {reason}. {Explanation of why this is intentional or not applicable}"
For each SKIPPED:
Run scripts/agent-reviews.js --reply <comment_id> "Skipped per user request"
DO NOT start Phase 2 until all replies are posted.
Phase 2: POLL FOR NEW COMMENTS (loop until quiet)
The watcher exits immediately when new comments are found (after a 5s grace period to catch batch posts). This means you run it in a loop: start watcher, process any comments it returns, restart watcher, repeat until the watcher times out with no new comments.
Step 6: Start Watcher Loop
Repeat the following until the watcher exits with no new comments:
6a. Launch the watcher in the background:
Run scripts/agent-reviews.js --watch --bots-only as a background task.
6b. Use TaskOutput to wait for the watcher to complete (blocks up to 12 minutes).
6c. Check the output:
- If new comments were found (output contains
EXITING WITH NEW COMMENTS):- Use
--detail <id>to read each new comment's full detail - Process them exactly as in Phase 1, Steps 3-5 (evaluate, fix, commit, push, reply)
- Go back to Step 6a to restart the watcher
- Use
- If no new comments (output contains
WATCH COMPLETE): Stop looping and move to the Summary Report.
Summary Report
After both phases complete, provide a summary:
## PR Review Bot Resolution Summary
### Results
- Fixed: X bugs
- Already fixed: X bugs
- Won't fix (false positives): X
- Skipped per user: X
### By Bot
#### cursor[bot]
- BUG-001: {description} - Fixed in {commit}
- BUG-002: {description} - Won't fix: {reason}
#### Copilot
- {description} - Fixed in {commit}
### Status
✅ All findings addressed. Watch completed.Important Notes
Response Policy
- Every finding gets a response - No silent ignores
- Responses help train bots and document decisions
- "Won't fix" responses prevent the same false positive from being re-raised
User Interaction
- Use
AskUserQuestionwhen uncertain about a finding - Don't guess on architectural or business logic questions
- It's better to ask than to make a wrong fix or wrong dismissal
Best Practices
- Verify findings before fixing - bots have false positives
- Keep fixes minimal and focused - don't refactor unrelated code
- Ensure type-check and lint pass before committing
- Group related fixes into a single commit
- Copilot
suggestionblocks often contain ready-to-use fixes
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
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.

