The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions define its role as a code pattern analyst. No bypass or override markers were found in the skill body or metadata. Role-play and role-setting instructions are standard and do not attempt to bypass safety protocols.\n- [COMMAND_EXECUTION]: The skill uses local Python and Bash scripts, such as pattern-detector.py, to automate analysis tasks. These scripts are provided as part of the skill's resource package and are executed locally using standard interpreters.\n- [EXTERNAL_DOWNLOADS]: No remote URLs or external dependency installations were detected. All resources, including reference catalogs and scripts, are stored in the local {baseDir} directory and accessed via local file paths.\n- [DATA_EXFILTRATION]: The skill reads local codebase files using Read and Grep tools for analysis. All data processing occurs within the agent's environment, and no patterns for network communication or data exfiltration were identified.\n- [SAFE]: The skill's behavior and implementation are consistent with its stated purpose of architectural and design pattern analysis. No malicious obfuscation, hidden instructions, or security vulnerabilities were found.

analyzing-patterns