Do pullrequest review

Receiving Code Review

Code review feedback requires technical evaluation, not automatic agreement. Verify before implementing. Push back when the feedback is wrong.

Preflight (mandatory)

Run this first. If it exits non-zero, halt and report the stderr output to the user verbatim. Do not proceed.

uv run maverick preflight do-pullrequest-review

Skip the preflight when invoked from inside a do-issue-* or do-epic phase — the orchestrator already ran a stricter one this session. Run it only when this skill is invoked standalone.

The check verifies the project is initialised and gh/uv are on PATH.

The Response Pattern

digraph receive_review {
    "Receive feedback" [shape=box];
    "Read all items completely" [shape=box];
    "All items understood?" [shape=diamond];
    "Ask for clarification on unclear items" [shape=box];
    "Verify each item against codebase" [shape=box];
    "Item technically correct?" [shape=diamond];
    "Implement fix" [shape=box];
    "Push back with reasoning" [shape=box];
    "Run verification after each fix" [shape=box];
    "More items?" [shape=diamond];
    "Run full verification suite" [shape=box];

    "Receive feedback" -> "Read all items completely";
    "Read all items completely" -> "All items understood?";
    "All items understood?" -> "Verify each item against codebase" [label="yes"];
    "All items understood?" -> "Ask for clarification on unclear items" [label="no"];
    "Ask for clarification on unclear items" -> "Verify each item against codebase";
    "Verify each item against codebase" -> "Item technically correct?";
    "Item technically correct?" -> "Implement fix" [label="yes"];
    "Item technically correct?" -> "Push back with reasoning" [label="no"];
    "Implement fix" -> "Run verification after each fix";
    "Run verification after each fix" -> "More items?";
    "Push back with reasoning" -> "More items?";
    "More items?" -> "Verify each item against codebase" [label="yes"];
    "More items?" -> "Run full verification suite" [label="no"];
}

Processing Review Feedback

1. Read All Items First

Read the complete review before acting on any single item. Items may be related — implementing one in isolation could conflict with another.

2. Clarify Before Implementing

If any item is unclear, ask for clarification on ALL unclear items before implementing ANY items.

Understand items 1, 2, 3, 6. Need clarification on items 4 and 5 before proceeding.

Do not implement the items you understand while waiting for clarification — partial understanding leads to partial (wrong) fixes.

3. Verify Against the Codebase

Before implementing each suggestion, verify it is technically correct for this codebase:

• Does the suggestion account for the full context? (The reviewer may not have seen all relevant code)
• Would implementing it break existing functionality?
• Is there a reason the current implementation is the way it is?
• Does it conflict with project conventions or architectural decisions?

4. Push Back When Wrong

Push back with technical reasoning when a suggestion is:

• Technically incorrect for this codebase
• Based on incomplete context (reviewer did not see related code)
• Over-engineering (adding abstraction for hypothetical requirements)
• Contradicting existing project patterns or conventions
• Violating YAGNI (suggesting features that are not needed)

How to push back:

Item 3: Reviewer suggests extracting a utility class.
Checked the codebase — this logic is used in one place only.
Extracting creates indirection without reuse benefit.
Keeping inline per YAGNI.

Do not push back defensively. State the technical facts.

5. Implement One at a Time

Process items in priority order:

1. Critical issues — must fix, blocks merge
2. Important issues — should fix before merge
3. Minor issues — fix if straightforward, note for later if not

After each fix, run verification (per the mav-local-verification skill). Do not batch multiple fixes without verifying between them.

Handling Review Sources

From the Code Reviewer Agent

The agent-code-reviewer agent provides structured two-stage feedback (spec compliance + code quality). Process it as follows:

• Spec compliance failures — these are requirements gaps. Fix them before anything else.
• Code quality criticals — must fix before merge
• Code quality importants — should fix
• Code quality minors — evaluate individually

From Human Reviewers (PR comments)

Human reviewers may have context the agent does not. However:

• Verify suggestions against the codebase before implementing
• If a suggestion conflicts with the project's established patterns, note this
• If a suggestion conflicts with the issue's requirements, flag it
• Reply in the PR comment thread, not as a top-level comment:

  gh api repos/{owner}/{repo}/pulls/{pr}/comments/{id}/replies \
    -f body="Fixed — extracted validation to boundary layer."
  

From CI/Automated Checks

CI failures are not suggestions — they are facts. Fix them. Use the mav-local-verification and mav-bp-cicd skills to diagnose and resolve.

What Not to Do

Anti-pattern	Why it fails	Instead
Implement everything without checking	Incorrect suggestions break code	Verify each item first
Implement some items, ask about others	Partial understanding leads to wrong fixes	Clarify all unclear items first
Ignore review feedback	Issues accumulate and compound	Process every item
Add extra improvements while fixing	Scope creep, muddies the review cycle	Fix only what was flagged
Batch all fixes, verify once at the end	Cannot isolate which fix broke something	Verify after each fix

Integration

When receiving review during a do-issue workflow:

1. Process the review per this skill
2. Implement fixes
3. Run verification per the mav-local-verification skill
4. If the review came from the agent-code-reviewer agent and had spec compliance failures, request a re-review after fixing
5. Update the plan comment on the issue if fixes changed the implementation approach

<!-- maverick-plugin-version: 4.0.1 -->