# Spec Kitty Review

> Perform structured code review and kanban transitions for completed task

- Skill: `richfrem/spec-kitty-review` (Agent Skill, multi-file: 20 files)
- Install (CLI): `npx skillmds@latest add richfrem/spec-kitty-review`
- Raw SKILL.md: https://api.skillmd.com/api/skills/richfrem/spec-kitty-review/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Productivity
- Author: richfrem (https://skillmd.com/u/richfrem)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/richfrem/spec-kitty-review

---


**IMPORTANT**: After running the command below, you'll see a LONG work package prompt (~1000+ lines).

**You MUST scroll to the BOTTOM** to see the completion commands!

Run this command to get the work package prompt and review instructions:

```bash
spec-kitty agent workflow review $ARGUMENTS --agent <your-name>
```

**CRITICAL**: You MUST provide `--agent <your-name>` to track who is reviewing!

If no WP ID is provided, it will automatically find the first work package with `lane: "for_review"` and move it to "doing" for you.

## Dependency checks (required)

- dependency_check: If the WP frontmatter lists `dependencies`, confirm each dependency WP is merged to the target branch before you review this WP.
- dependent_check: Identify any WPs that list this WP as a dependency and note their current lanes.
- rebase_warning: If you request changes AND any dependents exist, warn those agents to rebase and provide a concrete command (example: `cd .worktrees/FEATURE-WP02 && git rebase FEATURE-WP01`).
- verify_instruction: Confirm dependency declarations match actual code coupling (imports, shared modules, API contracts).

**After reviewing, scroll to the bottom and run ONE of these commands**:
- ✅ Approve: `spec-kitty agent tasks move-task WP## --to done --note "Review passed: <summary>"`
- ❌ Reject: Write feedback to the temp file path shown in the prompt, then run `spec-kitty agent tasks move-task WP## --to planned --review-feedback-file <temp-file-path>`

**The prompt will provide a unique temp file path for feedback - use that exact path to avoid conflicts with other agents!**

**The Python script handles all file updates automatically - no manual editing required!**