The Agent Skills Directory

[SAFE]: The skill provides methodological guidelines for agents to avoid false negatives when searching for files or code. It promotes rigorous verification and control-testing of search probes to ensure reliability.- [COMMAND_EXECUTION]: The skill mentions command-line tools like find and rg (ripgrep) within documentation blocks. These are used as benign examples of how an agent should perform searches and verify the integrity of their own tools.- [DATA_EXFILTRATION]: No network operations or commands to send data to external servers were detected. References to local file paths are used exclusively for instructional search examples.- [PROMPT_INJECTION]: No malicious instruction overrides, role-play injections, or safety bypass attempts were detected. The content focuses on diagnostic accuracy.

proving-an-absence