The Agent Skills Directory

[SAFE]: The skill provides documentation and patterns for regression testing in AI-assisted development environments. Analysis across all 10 threat categories revealed no malicious intent or security vulnerabilities.- [CREDENTIALS_SAFE]: The provided code snippets for test setup demonstrate safe practices by explicitly clearing sensitive environment variables (e.g., Supabase URL and keys) to ensure tests run in a mock/sandbox environment without interacting with real production databases.- [COMMAND_EXECUTION]: The suggested workflow involves standard development commands such as npm run test and npm run build. These are used appropriately for their intended purpose of verifying code changes and catching type errors.- [INDIRECT_PROMPT_INJECTION]: The skill identifies the risk of AI 'blind spots' where the model reviews its own code and suggests automated testing as a mitigation. It correctly categorizes potential failure modes without introducing instructions that could be exploited to bypass safety filters.

ai-regression-testing