The Agent Skills Directory

[SAFE]: The skill provides a comprehensive framework for self-verification, ensuring that work is thoroughly tested and verified before being reported as complete. It emphasizes evidence-based claims and follows best practices for software development verification workflows.- [COMMAND_EXECUTION]: The skill utilizes Bash and grep to perform local verification tasks, such as searching for code stubs, checking for incomplete implementations (e.g., TODOs or empty returns), and running local build/test scripts (e.g., npm test, npm run build). These actions are performed using tools explicitly allowed in the skill's configuration and are appropriate for its stated purpose.- [PROMPT_INJECTION]: No evidence of prompt injection or attempts to bypass security constraints were found. The instructional language is professional and directed at improving work quality and verification rigor, using structured templates to ensure consistency in reporting.

verification-before-completion