The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands like grep, find, and npx madge as defined in references/analysis-techniques.md to perform dependency tracking and pattern analysis. These commands are used to read and process the codebase to generate risk assessments and reports.
[EXTERNAL_DOWNLOADS]: The use of npx madge in references/analysis-techniques.md may trigger a download of the madge package and its dependencies from the npm registry. As npm is a well-known service, this is a standard development operation.
[PROMPT_INJECTION]: The skill analyzes project source code which could contain malicious instructions. Ingestion points: Project source files analyzed via grep and madge (referenced in references/analysis-techniques.md). Boundary markers: Not present. Capability inventory: Restricted to read-only shell commands and image generation (e.g., impact.svg) defined in references/analysis-techniques.md, with no network exfiltration or code modification capabilities. Sanitization: No explicit sanitization of input file content is performed.

Ripple