code-review
Code Review
Dispatch specialized review agents matched to the situation. Choose a mode or let context guide you.
Review Modes
Quick Review
When: Ad-hoc checks, quick sanity check, low-noise feedback Dispatch: quick-reviewer agent (confidence scoring, reports only >= 80) Output: High-confidence issues grouped by severity (Critical 90-100, Important 80-89)
Deep Review
When: After completing a feature, before merge, in plan-execution review loop Dispatch: deep-reviewer agent (full checklist with template placeholders) Placeholders: {WHAT_WAS_IMPLEMENTED}, {PLAN_OR_REQUIREMENTS}, {BASE_SHA}, {HEAD_SHA}, {DESCRIPTION} Output: Strengths, Issues (Critical/Important/Minor), Recommendations, Merge verdict
Specialist Review
When: Targeted analysis of a specific quality dimension Dispatch: One or more specialist agents:
| Agent | Focus |
|---|---|
| comment-analyzer | Comment accuracy, documentation quality, comment rot |
| test-analyzer | Test coverage quality, behavioral coverage gaps, test resilience |
| silent-failure-hunter | Error handling, silent failures, catch block specificity |
| type-design-analyzer | Type invariants, encapsulation, type usefulness |
Comprehensive Review
When: Before merging to main, finishing a development branch Dispatch: 4-pass structure (TDD compliance, diff-only quality, contextual quality, browser audit). See references/comprehensive-review-procedure.md for pass details and ordering.
How to Request
- Gather git context using
scripts/review-prep(auto-detects base branch, outputs BASE_SHA, HEAD_SHA, diff stats, changed files, commit log). Use--jsonfor structured output. - Read the reference file for your chosen agent(s)
- Spawn
general-purposeTask agent with the prompt + review-prep output - Act on feedback: fix Critical immediately, fix Important before proceeding
Workflow Integration
| Workflow | Review Mode |
|---|---|
| plan-execution (per-task) | Deep Review |
| plan-execution (batch) | Deep Review |
| Before merge to main | Comprehensive Review |
| Ad-hoc / when stuck | Quick Review |
| Specific concern | Specialist Review |
| GitHub PR review + posting | github-pr skill (automated PR review orchestrator) |
Simplification Mode
After review, optionally dispatch a simplification pass to improve clarity, consistency, and maintainability without changing functionality. Read references/simplifier.md for the full agent prompt, then spawn a general-purpose Task agent with that prompt targeting the reviewed files.
Handling Feedback
When receiving code review feedback (from humans or other agents), follow the structured 6-step response pattern: READ, UNDERSTAND, VERIFY, EVALUATE, RESPOND, IMPLEMENT. No performative agreement — verify before implementing, push back with technical reasoning when warranted. See references/handling-feedback.md for the full pattern, push-back criteria, YAGNI checks, and source-specific handling.
Red Flags
- Skipping review because "it's simple"
- Ignoring Critical issues
- Proceeding with unfixed Important issues
- Running Comprehensive passes all sequentially (1+2 are independent, parallelize)
- Not deduplicating between Pass 2 and Pass 3 in Comprehensive mode
Reference Files
| Agent | Reference |
|---|---|
| deep-reviewer | references/deep-reviewer.md |
| quick-reviewer | references/quick-reviewer.md |
| comment-analyzer | references/comment-analyzer.md |
| test-analyzer | references/test-analyzer.md |
| silent-failure-hunter | references/silent-failure-hunter.md |
| type-design-analyzer | references/type-design-analyzer.md |
| simplifier | references/simplifier.md |
| handling-feedback | references/handling-feedback.md |
More from hjewkes/agent-skills
self-improve
Use when a session produced reusable insights, when the user says "learn from this", "remember this", or "improve yourself", or after completing a complex task where patterns were discovered
64md-render
Use when asked to render, preview, or view a markdown file in the browser. Triggers on "render markdown", "preview this", "show me this document", "open in browser".
21skills-management
Use when creating, finding, installing, reviewing, or managing Claude Code skills — covers skill authoring, discovery, conventions, and lifecycle management
14github-pr
GitHub PR workflow — creating PRs, posting automated review comments, managing PR feedback cycles. Use when code is reviewed and ready for GitHub.
14buildkite
Buildkite CI/CD integration. Use when the user needs to check build status, trigger builds, read build logs, debug failures, manage pipelines, or any Buildkite workflow. Triggers include "buildkite", "build", "pipeline", "CI", "deploy", "build log", "build failed".
14agent-browser
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction.
13