extrabrain-prepare-interview
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands to manage temporary directories and runs a local Python script (validate_prep.py) for file validation. The directory creation uses restrictive permissions (-m 700), and the validation script only performs structural analysis of text files without executing their content.\n- [INDIRECT_PROMPT_INJECTION]: The skill handles untrusted data from user uploads and web search results.\n
- Ingestion points: User-provided interview descriptions, job postings, resumes, and web research results.\n
- Boundary markers: Includes a system instruction to treat input as data, but lacks explicit formatting delimiters.\n
- Capability inventory: File system access (mkdir), script execution (python3), and vendor CLI integration (extrabrain).\n
- Sanitization: Relies on internal AI guardrails and the 'treat as data' instruction.\n- [EXTERNAL_DOWNLOADS]: The skill performs web searches to gather company and role information. These operations are conducted through authorized platform tools and are essential to the skill's primary function.
Audit Metadata