skills/duthaho/skillhub/jobfit/Gen Agent Trust Hub

jobfit

Pass

Audited by Gen Agent Trust Hub on Jul 25, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted data fetched from external URLs such as job postings and user CVs. This creates an indirect prompt injection surface where malicious content in these external sources could attempt to influence the agent's logic.
  • Ingestion points: External content enters the context via WebFetch and WebSearch operations defined in SKILL.md (Step 0, Step 0b, and Step 1).
  • Boundary markers: The skill lacks explicit prompt delimiters or 'ignore' instructions for external content, relying instead on high-level instructions to 'extract only what's actually stated'.
  • Capability inventory: The agent can write files, execute simple shell commands, and make network requests.
  • Sanitization: There are explicit instructions for the agent to avoid fabrication and remain grounded in the source text.
  • [COMMAND_EXECUTION]: The skill executes a localized shell command (date +%F) to generate a timestamp for output files. This is a low-risk, benign operation.
  • [EXTERNAL_DOWNLOADS]: The skill uses WebFetch and WebSearch to retrieve information from reputable job board platforms and professional networking sites (e.g., LinkedIn, Greenhouse, Lever) as part of its core functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 25, 2026, 04:21 PM
Security Audit — agent-trust-hub — jobfit