job-match-finder
Pass
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill navigates to and extracts data from LinkedIn (linkedin.com), a well-known service, to perform its job search function.
- [COMMAND_EXECUTION]: The skill instructions suggest using the
pdftotextutility to process PDF files if a native skill is unavailable, which is a standard procedure for document analysis. - [PROMPT_INJECTION]: The skill has an inherent attack surface for indirect prompt injection because it ingests untrusted data from external sources.
- Ingestion points: LinkedIn profile pages, job listing descriptions, and user-uploaded resumes (PDF/DOCX).
- Boundary markers: None identified. The instructions do not explicitly tell the agent to ignore instructions embedded in the external text.
- Capability inventory: Browser control via Claude in Chrome, and file system write access for report generation.
- Sanitization: The skill recommends stripping URLs from extracted text to avoid platform privacy filters, but does not specify measures to sanitize text against injection or XSS within the generated HTML report.
Audit Metadata