The Agent Skills Directory

[PROMPT_INJECTION]: The skill displays a surface for indirect prompt injection (Category 8) because it recursively reads and processes the contents of files within a codebase to generate summaries. 1. Ingestion points: Step 3 involves reading all files in a directory to analyze purpose and relationships. 2. Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from obeying instructions embedded within the analyzed files. 3. Capability inventory: The skill has the ability to read and write files and execute local shell commands. 4. Sanitization: No sanitization or filtering of codebase content is mentioned before it is passed to the generation agent.
[COMMAND_EXECUTION]: The skill uses local shell commands, specifically 'find' and 'grep', in Step 5 to validate the hierarchy and parent references of the generated documentation. These are standard utility operations.

deepinit