Agent skill

exploration-review

Interactive code review through conversation. HOUSTON guides review, spawns specialized agents, and helps create Beads for issues found.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/exploration-review

SKILL.md

/exploration-review - Interactive Code Review

Review code through conversation. This is collaborative analysis, not a report dump. You guide the review, ask questions, spawn specialized agents, and work through findings together.

The Process

  1. Understand scope - What code to review? Recent changes, specific files, or feature area?
  2. Ask which categories - Quality, security, performance, simplification, or all?
  3. Spawn relevant agents - Run in background while you talk
  4. Work through findings - Discuss issues, get context, prioritize
  5. Create report - Summarize findings by priority
  6. Offer Beads - Ask if user wants to track issues

Starting the Review

1. Determine Scope

Ask what to review:

  • Recent changes (git diff)
  • Specific files or directories
  • A feature or component
  • Code from last /mission

2. Select Categories

Use AskUserQuestion:

"Which areas should I focus on?"
Options:
- Quality (readability, structure, patterns)
- Security (secrets, injection, validation)
- Performance (algorithms, queries, optimization)
- Simplification (dead code, over-engineering, DRY)
- All of the above

3. Spawn Agents

Based on selection, spawn with run_in_background: true:

Category Agent Focus
Quality space-agents:review-quality Readability, naming, complexity, patterns
Security space-agents:review-security Secrets, injection, auth, OWASP
Performance space-agents:review-performance Algorithms, queries, caching, bundle
Simplification space-agents:review-code-simplifier Dead code, over-engineering, DRY, bloat

Continue conversation while agents work. Check results with TaskOutput block: false.

Working Through Findings

Priority Levels

Categorize all findings:

Priority Meaning Action
Critical Security vulnerability, data loss risk, broken functionality Must fix before merge
Warning Code smell, maintainability issue, potential bug Should fix
Suggestion Style improvement, optimization opportunity Consider improving

Discussion Flow

For each finding:

  1. Present the issue - What, where, why it matters
  2. Get context - Ask if there's a reason for current approach
  3. Discuss fix - Agree on solution or accept as-is
  4. Categorize - Confirm priority level

Red flags to watch for:

  • User dismissing Critical issues - push back
  • "It works so it's fine" - explain long-term cost
  • Over-engineering suggestions - keep it practical

Your Role

  • Ask questions - Understand context before judging
  • Have opinions - Recommend priorities, push back on bad patterns
  • Suggest agents, don't auto-spawn - Always ask first
  • Be constructive - Acknowledge what's done well, not just problems
  • Keep talking - Never wait silently for agent results

Available Agents

Spawn with run_in_background: true, continue conversation immediately:

  • space-agents:review-quality - Code quality and maintainability
  • space-agents:review-security - Security vulnerabilities and risks
  • space-agents:review-performance - Performance issues and optimizations
  • space-agents:review-code-simplifier - Dead code, over-engineering, DRY violations

AskUserQuestion (Required)

Always use AskUserQuestion for every question in review. Prefer multiple choice when you can anticipate likely answers.

Output

When review is complete:

1. Summary Report

Present findings organized by priority:

## Review Summary

### Critical (must fix)
- [Issue with file:line reference]

### Warnings (should fix)
- [Issue with file:line reference]

### Suggestions (consider)
- [Issue with file:line reference]

### What's Good
- [Positive observations]

2. Offer Beads

Ask user:

AskUserQuestion:
  "Want to create Beads to track these issues?"
  Options:
  - "Yes, create bugs for Critical/Warning" - Track issues that need fixing
  - "Yes, create tasks for all" - Track everything including suggestions
  - "No, I'll handle it" - Skip Bead creation

If creating Beads:

  • Use bd create --type=bug for Critical/Warning issues
  • Use bd create --type=task for Suggestions
  • Include file:line references in description

Didn't find tool you were looking for?

Be as detailed as possible for better results