| name | claude-stop-review-hook |
| description | Configure or explain the optional Codex Stop hook that can run a read-only Claude review when a turn ends. |
Claude Stop Review Hook
Use this skill when the user wants Claude Code to review changes automatically when a Codex turn stops.
The plugin bundles hooks/hooks.json with a Stop hook, but the hook helper is inert until explicitly configured with the plugin root and then enabled for review. Codex also requires non-managed hooks to be reviewed and trusted with /hooks before they run.
First point the bundled hook config at this plugin using an absolute path before launching Codex:
export CLAUDE_COMPANION_PLUGIN_ROOT=/absolute/path/to/cc-plugin-codex
Then enable review in the shell before launching Codex:
export CLAUDE_COMPANION_STOP_REVIEW=1
Or enable it per repository:
mkdir -p .codex
touch .codex/claude-stop-review.enabled
The per-repository marker enables review after the hook helper is located; it does not replace CLAUDE_COMPANION_PLUGIN_ROOT.
By default, the hook reports review findings without blocking Codex. To block on high or critical findings:
export CLAUDE_COMPANION_STOP_REVIEW_BLOCKING=1
Useful options:
CLAUDE_COMPANION_STOP_REVIEW_MODEL=opus to choose the model for serious review gates.CLAUDE_COMPANION_STOP_REVIEW_EFFORT=low to choose Claude Code effort.CLAUDE_COMPANION_STOP_REVIEW_BASE=main to review main...HEAD.CLAUDE_COMPANION_STOP_REVIEW_ADVERSARIAL=1 to use the adversarial review prompt.
Model guidance:
- Use
CLAUDE_COMPANION_STOP_REVIEW_MODEL=opus for serious blocking gates or important merge-readiness checks.
- Use
CLAUDE_COMPANION_STOP_REVIEW_EFFORT=low only for smoke tests, cheap sanity checks, or explicitly low-effort requests.
- Do not assume short Claude model names always map to the expected backend model. After real Claude calls, inspect and report actual model usage from JSON output when available.
Manual smoke test:
Use the companion script from the installed plugin root. If the current working
directory is not this plugin checkout, resolve the script relative to this skill
file, for example ../../scripts/claude-companion.mjs.
node scripts/claude-companion.mjs hook-stop-review --json
The hook always uses Claude Code --permission-mode plan and should not edit files.
