The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use the rg (ripgrep) utility to trace references across the codebase.
[PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill reads and interprets untrusted content from the codebase to guide its deletion and refactoring logic. * Ingestion points: The skill analyzes all project files, including source code, comments, and documentation, during the 'Sweep' phase. * Boundary markers: Absent. The agent processes file content directly without delimiters or safety warnings to ignore embedded instructions. * Capability inventory: The skill has the authority to delete files and modify code across the entire project. * Sanitization: None. There is no validation or filtering of the content being read before the agent acts upon it.

hard-cut