Validation Phase
Validates implementation quality through tiered build checks, automated acceptance criteria verification, and agent-driven code reviews.
The source repository doesn't declare a license. Check its terms before reusing the code.
Key features
- Intelligent failure handling with critical blocking for build and requirement gaps
- Tiered validation scaling (Build, Lint, Prettier, Tests) based on change volume
- Multi-agent orchestration for deep code quality and security analysis
- Weighted 5-point confidence assessment for complex architectural changes
- Automated Acceptance Criteria (AC) verification against PRD documentation
Use cases
- Ensuring large-scale refactors maintain architectural integrity and pass all tests
- Enforcing project-specific quality standards before finalizing a code change
- Automating the verification of 'Definition of Done' against requirement documents
FAQ
When should I use this skill in my workflow?
This skill is designed to run automatically after the execution phase of commands like /audit, /build, or /ms. It acts as a quality gate to catch errors before an AI agent reports a task as complete.
What happens if my project fails the validation checks?
The skill provides intelligent failure handling. Build failures and incomplete acceptance criteria result in a 'CRITICAL' block that prevents the process from continuing. For linting or test issues, it provides actionable recommendations or asks for user guidance.
What does the Validation Phase skill do?
The Validation Phase skill automates the verification of code implementations by running tiered build checks, verifying completion of acceptance criteria in your PRD, and orchestrating multi-agent code reviews to ensure production readiness.
How does this skill improve AI-driven development quality?
It introduces a weighted 5-point confidence assessment for complex tasks and utilizes specialized sub-agents to perform deep security and architectural analysis, significantly reducing technical debt and hallucinations.
How does the tiered validation scaling work?
To optimize performance, the skill scales checks based on change size: 'Small' changes run a build only, 'Medium' changes add linting and formatting, and 'Large' changes (11+ files) trigger the full suite, including tests.
Related skills
Code Review
Analyzes code for security vulnerabilities, performance bottlenecks, and adherence to Rust-specific best practices to ensure high-quality, maintainable software.
Security Testing29712 ptsFrontend Code Review
Conducts comprehensive audits of React, TypeScript, and Tailwind codebases to identify security vulnerabilities, performance bottlenecks, and accessibility gaps.
Security Testing11212 ptsUnified Review Orchestration
Orchestrates and runs appropriate code review skills based on codebase analysis and user context.
Security Testing212 pts