Agent skill
review-codex
Final code/plan review using Codex CLI. Use as the last review step after sonnet.
Install this agent skill to your Project
npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/review-codex
SKILL.md
Review Codex (Final Review)
You are the final reviewer, invoking Codex CLI for the ultimate review before approval.
Reference Documents
First, read and understand:
skill/multi-ai/reference/standards.md- Review criteria and decision rules
Setup: Copy Schemas to Task Directory
Before invoking Codex, copy the required schema files to .task/ so Codex can access them:
mkdir -p .task
cp skill/multi-ai/reference/schemas/plan-review.schema.json .task/
cp skill/multi-ai/reference/schemas/review-result.schema.json .task/
Your Role
- Final gate: Last check before plan approval or code completion
- External review: Use Codex CLI for independent assessment
- Structured output: Enforce JSON schema for consistent results
Determine Review Type
Check which files exist:
- If
.task/plan-refined.jsonexists and no.task/impl-result.json-> Plan Review - If
.task/impl-result.jsonexists -> Code Review
Session Management
Check if .task/.codex-session-active exists:
- If yes: This is a subsequent review (Codex has reviewed before)
- If no: This is a first review
For Plan Reviews
- Read
.task/plan-refined.json - Invoke Codex using Bash:
codex exec \
--full-auto \
--output-schema ".task/plan-review.schema.json" \
-o .task/review-codex.json \
"Review the plan in .task/plan-refined.json. Check for completeness, feasibility, security concerns, and potential issues. Apply standards from skill/multi-ai/reference/standards.md."
For Code Reviews
- Read
.task/impl-result.json - Invoke Codex using Bash:
codex exec \
--full-auto \
--output-schema ".task/review-result.schema.json" \
-o .task/review-codex.json \
"Review the implementation in .task/impl-result.json. Identify bugs, security issues, code style violations. Apply standards from skill/multi-ai/reference/standards.md."
For Subsequent Reviews
If .task/.codex-session-active exists, use resume:
codex exec \
--full-auto \
--output-schema ".task/review-result.schema.json" \
-o .task/review-codex.json \
resume --last \
"Re-review the changes. Previous issues should be addressed."
After Codex Completes
- Mark session as active:
touch .task/.codex-session-active - Read
.task/review-codex.jsonto get the result - Report back:
- Review type (plan or code)
- Status from Codex (approved or needs_changes)
- Summary of Codex findings
- Confirm output in
.task/review-codex.json
If Codex Fails
If the Bash command fails or output is invalid:
- Report the error to the user
- Check if
codexCLI is installed and authenticated - Try the command again with verbose output
- If persistent failure, ask user to verify Codex CLI setup
Recommended Agent Skills
Expand your agent's capabilities with these related and highly-rated skills.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
agent-ops-spec
Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-testing
Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.
agent-ops-state
Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.
Didn't find tool you were looking for?