The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/parse_resume.py file executes the mineru command-line utility to convert PDF documents to Markdown. The process uses subprocess.run with an argument list, which is a secure method for executing external commands as it prevents shell injection by not invoking a shell to interpret the arguments.
[EXTERNAL_DOWNLOADS]: The skill interfaces with the MinerU Agent API at mineru.net to parse documents. This process involves uploading document content to the service and downloading the processed Markdown result. This is a legitimate functional dependency for the skill's resume-parsing features.
[INDIRECT_PROMPT_INJECTION]: The skill analysis surface involves processing external resume data and job descriptions.
Ingestion points: Resume data is processed in scripts/parse_resume.py (files/URLs), and job descriptions are ingested through scripts/interview_session.py.
Boundary markers: Prompt templates in references/ use explicit section delimiters and JSON structures to maintain separation between instructions and external content.
Capability inventory: The skill possesses capabilities for local script execution, file persistence, and network requests to its document processing service.
Sanitization: Content is validated against expected JSON schemas before being re-introduced to the LLM or saved to the session state.

cs-tech-interviewer