What happens if I don't have the Gemini or Codex CLIs configured?

The skill includes a graceful fallback system. If external CLIs are missing or unauthenticated, it automatically launches Claude Task agents to fill those review slots, ensuring you still receive a complete 6-perspective report.

How does the skill decide if a code change passes or fails?

It uses a conservative synthesis logic. Any critical security concern or failed behavioral criterion results in a 'Revise' verdict, while moderate issues lead to a 'Conditional Pass' to ensure high standards.

Does this skill provide numerical scores for my code?

No, the skill provides narrative verdicts (Pass, Conditional Pass, or Revise) and specific suggestions. This avoids arbitrary scoring and focuses on actionable feedback for the developer.

How are the domain experts chosen for the review?

The skill resolves experts from an internal roster based on the project context, ensuring the feedback is relevant to the specific framework or language being used.

Can I use this skill to review project plans before writing code?

Yes, by using the '--mode plan' flag, you can trigger a specialized review of your implementation strategy and technical specifications to catch logic errors before coding begins.

CWF Review: Multi-Agent Quality Gates for Claude Code

The CWF Review skill provides a standardized, multi-layered quality assurance framework for Claude Code, utilizing six parallel review agents to evaluate development artifacts. It operates across three distinct stages—requirement clarification, implementation planning, and final code review—ensuring consistent quality gates throughout the software development lifecycle. By combining internal security and UX audits with external perspectives from Gemini or Codex and two specialized domain experts, it offers a diverse and comprehensive feedback loop that catches edge cases, architectural flaws, and security vulnerabilities before they reach production.

주요 기능

01External provider integration with Google Gemini and OpenAI Codex for diverse model perspectives.

02Automated session-log cross-checking to ensure code changes align with development history.

0373 GitHub stars

04Six parallel review slots including specialized agents for Security, UX/DX, and Domain Expertise.

05Robust fallback mechanism that transitions to Claude Task agents if external CLIs are unavailable.

06Support for three modes: 'clarify' for requirements, 'plan' for specifications, and 'code' for implementations.

사용 사례

01Validating project requirements against holdout scenarios to ensure feature parity and correctness.

02Reviewing technical specifications and implementation plans to identify architectural risks early.

03Conducting a comprehensive security and UX audit before shipping code to production.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add corca-ai/claude-plugins review

For use in Claude.ai and ChatGPT

Download Skill

주요 기능

01External provider integration with Google Gemini and OpenAI Codex for diverse model perspectives.

02Automated session-log cross-checking to ensure code changes align with development history.

0373 GitHub stars

04Six parallel review slots including specialized agents for Security, UX/DX, and Domain Expertise.

05Robust fallback mechanism that transitions to Claude Task agents if external CLIs are unavailable.

06Support for three modes: 'clarify' for requirements, 'plan' for specifications, and 'code' for implementations.

사용 사례

01Validating project requirements against holdout scenarios to ensure feature parity and correctness.

02Reviewing technical specifications and implementation plans to identify architectural risks early.

03Conducting a comprehensive security and UX audit before shipping code to production.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add corca-ai/claude-plugins review

For use in Claude.ai and ChatGPT

Download Skill

CWF Multi-Perspective Reviewer

CWF Multi-Perspective Reviewer

주요 기능

사용 사례

주요 기능

사용 사례