Claude Plan Reviewer — AI skill for Claude Code
Automatically review Claude Code plans using external AI CLIs.
How to install Claude Plan Reviewer
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open yuuichieguchi/claude-plan-reviewer and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Claude Plan Reviewer does
Automatically review Claude Code plans using external AI CLIs.
Alternatives in AI
- Flow Deliver — Multi-AI validation, scoring, and review using Codex and Gemini CLIs (Double Diamond Deliver phase) 2.8k ★
- Claudish CLI — for external AI models) 679 ★
- Fusion Plan — OMC interview (interactive) → 3-round Opus 4.8 + GPT-5.5 fusion deepening → concise .omc/plans/ output → revie 468 ★
README
claude-plan-reviewer
Automatically review Claude Code plans using external AI CLIs (Codex, Gemini CLI). Uses Claude Code's `PreToolUse` hook to intercept `ExitPlanMode`, send the plan for external review, and inject feedback back into Claude's context.
How It Works
Claude writes a plan and calls ExitPlanMode
→ PreToolUse hook fires → hook runs
→ Finds the latest plan file from ~/.claude/plans/
→ Runs review via Codex / Gemini CLI
→ If LGTM → permissionDecision:"allow" → Claude exits plan mode
→ If not → permissionDecision:"deny" + review feedback
→ Claude revises the plan → calls ExitPlanMode again
→ After maxReviews reached → allow without review
**Note:** Due to a Claude Code limitation, the review result is only displayed in the chat when the reviewer returns feedback (deny). When the reviewer returns LGTM (allow), the review result is streamed to stderr but not shown in the chat UI.
Install
npm install -g claude-plan-reviewer
Setup
claude-plan-reviewer setup
**Note:** The hook takes effect in Claude Code sessions started after setup. Already-running sessions are not affected.
Teardown
claude-plan-reviewer teardown
Uninstall
npm uninstall -g claude-plan-reviewer
Configuration
# Show current config
claude-plan-reviewer config show
# Change adapter (codex | gemini)
claude-plan-reviewer config set adapter gemini
# Change max reviews per session
claude-plan-reviewer config set maxReviews 3
# Set Codex model
claude-plan-reviewer config set codex.model o3
# Set Gemini model
claude-plan-reviewer config set gemini.model gemini-2.5-pro
Config File
`~/.claude-plan-reviewer.json`
{
"adapter": "codex",
"maxReviews": 2,
"prompt": "",
"codex": {
"model": "",
"sandbox": "read-only"
},
"gemini": {
"model": ""
}
}
| Key | Description | Default |
|---|---|---|
adapter |
Reviewer t |
Related Skills
Multipov Plan
Run a multi-agent design-doc review on $ARGUMENTS (path to a markdown design doc, typically docs/plans/YYYY-MM
Efficient Fable Orchestrator
Claude Code skill: turns a session into a quota-conscious orchestrator that plans at a high level and delegate
Dsh Auto Review
Second-model AI auto-review for DeepSeek Harness approval requests: a read-only reviewer subagent returns stru
Agy Plugin Cc
agy-plugin-cc is a Claude Code companion for running Gemini CLI or Antigravity CLI (agy) as a cross-model task
Claude Code Reviewer
An adversarial reviewer workflow for Claude Code that enforces mandatory AI review.
Gh Review Loop
Fixes the AI reviewer findings on your PR, including the sibling instances the bot missed. Runs your tests bef
Related Agents
Subprocess Safety Reviewer
Reviews changes that spawn or control external processes (tmux, git, gh, bash, agent CLIs) and changes touchin
Code Writer Go
Go code implementation sub-agent. Invoked by any manager whose plan calls for Go code — Kubernetes operators/c
External Pattern Checker
Opt-in bounded web research on plan claims prior-art-checker left Silent, on empirically-tricky topics. Sideca