Delegate independent tasks to specialized agents working concurrently with isolated context.
Cross-Review Policy Skill
cross-review-policy
Cross-Review Policy Skill
SKILL.md
Full skill instructions
Cross-Review Policy Skill
Version: 1.0.0 Category: Workflows Triggers: Completing AI work, code review, presenting to user
Quick Reference
Core Rule
ALL work performed by Claude Code or Google Gemini MUST be reviewed by OpenAI Codex.
Review Flow
Claude/Gemini performs task
↓
Commit changes
↓
Codex reviews
↓
Feedback needed?
├── No → Present to user (APPROVED)
└── Yes → Implement fixes
↓
Iteration < 3?
├── Yes → Re-commit → Codex reviews
└── No → Present to user (LIMIT_REACHED)
Maximum Iterations: 3
| Iteration | Status | Action |
|---|---|---|
| 1 | First review | Codex reviews original commit |
| 2 | Second review | Codex reviews fix commit |
| 3 | Final review | Last chance for approval |
| 3+ | Limit reached | Present to user regardless |
Commands
# Trigger cross-review loop
./scripts/ai-review/cross-review-loop.sh --max-iterations 3
# Check review status
./scripts/ai-review/review-manager.sh list
# Check iteration status
./scripts/ai-review/review-manager.sh iteration-status <review_id>
# Force complete (skip remaining iterations)
./scripts/ai-review/review-manager.sh force-complete <review_id>
Commit Message Formats
Original Work Commit
<type>: <description>
<body>
Co-Authored-By: Claude <[email protected]>
Review Fix Commit
fix(review): <description>
Addresses Codex review feedback (iteration N/3)
Review-ID: <review_id>
Original-Commit: <original_sha>
Co-Authored-By: Claude <[email protected]>
Review Types Covered
| Review Type | Trigger |
|---|---|
| Code Reviews | Every commit |
| SPARC Gate-Pass | Phase completion |
| QA Reviews | Feature completion |
| Security Reviews | Security-sensitive changes |
| Architecture Reviews | Architectural changes |
Provider Mapping
| Primary Agent | Reviewer |
|---|---|
| Claude Code | OpenAI Codex |
| Google Gemini | OpenAI Codex |
| OpenAI Codex | N/A (is the reviewer) |
Exit Conditions
Review loop exits when ANY condition is met:
- APPROVED: Codex approves the changes
- MAX_ITERATIONS: 3 iterations completed
- FORCE_COMPLETE: User forces completion
- NO_CHANGES: No actionable feedback
Exceptions (NOT Required)
- Documentation-only changes (pure markdown/text)
- Configuration files (non-code config)
- Emergency hotfixes (with explicit user override)
- User-initiated commits (non-AI commits)
Full Reference
See: @docs/modules/ai/CROSS_REVIEW_POLICY.md
Use this when completing AI work, understanding review requirements, or managing review iterations.
More Other skills
brand-guidelines
Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.
theme-factory
Toolkit for styling artifacts with a theme. These artifacts can be slides, docs, reportings, HTML landing pages, etc. There are 10 pre-set themes with colors/fonts that you can apply to any artifact that has been creating, or can generate a new theme on-the-fly.
internal-comms
A set of resources to help me write all kinds of internal communications, using the formats that my company likes to use. Claude should use this skill whenever asked to write some sort of internal communications (status reports, leadership updates, 3P updates, company newsletters, FAQs, incident reports, project updates, etc.).
Fundamental design principles for writing maintainable, testable Python code.
responsive-design
Modern responsive layouts using container queries, fluid typography, CSS Grid, and mobile-first strategies.
Generate Playwright tests from scenarios using interactive browser exploration and validation.
Automated website exploration and test case generation using Playwright.
Modern React state management with Redux Toolkit, Zustand, Jotai, and React Query for every state category.
gdpr-data-handling
GDPR-compliant data handling with consent management, data subject rights, and privacy controls.
Python error handling patterns including input validation, exception hierarchies, and partial failure handling. Use when implementing validation logic, designing exception strategies, handling batch processing failures, or building robust APIs.
refactor-plan
Detailed multi-file refactor planning with safety checks and rollback guidance.
