The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from external resume files and job descriptions, which could potentially contain hidden instructions.
Ingestion points: Reads from resume.yaml and user-provided job description (JD) paths.
Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands in the processed data.
Capability inventory: The skill possesses file read and write permissions within the knowledge/sessions/ directory.
Sanitization: No explicit content sanitization or validation of the input text is performed before processing.
[DYNAMIC_EXECUTION]: The skill dynamically loads agent instructions from local files to define persona-based behavior at runtime.
The skill reads definitions from local markdown files (e.g., agents/recruiter.agent.md, agents/engineer-peer.agent.md).
These instructions are adopted as the active persona for evaluating resumes, allowing for instruction-set modularity.

review