Can it automatically fix issues it finds?

Yes, when the trust level is set to medium, the skill will automatically loop back to the implementation phase to attempt fixes up to three times before escalating to the user.

Does it work with different programming languages?

Yes, it detects the primary language from the diff and applies language-specific supplements and red flags for a tailored review process.

How does the skill decide to reject a change?

It uses a scoring system where any critical concern, a stakeholder rejection, or three or more high-priority concerns result in an automatic rejection.

What stakeholders are included in the review?

The skill includes specialized agents for architecture, security, quality, testing, performance, UX, sales, and marketing, ensuring comprehensive coverage.

Stakeholder Review Gate

Name: Stakeholder Review Gate
Author: cowwoc

bycowwoc

•

Security & Testing

Automates multi-perspective quality assurance reviews using specialized AI subagents for architecture, security, performance, and business alignment.

The Stakeholder Review skill acts as an automated quality gate within the Claude Code workflow, ensuring that implementation changes are scrutinized from multiple professional perspectives before final approval. It spawns specialized subagents representing roles like Security Architects, Performance Engineers, and QA Testers to evaluate code diffs against specific domain criteria. By aggregating these cross-functional insights and applying severity-based decision rules, the skill helps developers catch critical issues, maintain high code standards, and ensure business alignment without manual oversight.

Key Features

01Language-specific pattern detection and red-flagging

02Detailed review reporting with actionable recommendations

03Configurable trust levels for automatic fix-and-loop iterations

04Automated severity-based approval and rejection logic

052 GitHub stars

06Parallel multi-role reviews (Security, Performance, UX, Architecture, etc.)

Use Cases

01Automating security and performance audits during the active development cycle

02Validating complex architectural changes before merging to main branch

03Ensuring marketing and sales alignment for user-facing feature changes

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add cowwoc/claude-code-cat stakeholder-review

For use in Claude.ai and ChatGPT

Download Skill