blog-geo

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides an optional feature to query Google Search Console data by executing a local Python script (skills/blog-google/scripts/run.py). This is a functional requirement for providing search performance context and is restricted to the vendor's own script ecosystem.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection as it processes untrusted blog content during the audit.
  • Ingestion points: Content is extracted from a user-provided file in SKILL.md (Step 1).
  • Boundary markers: The instructions do not define explicit delimiters to separate user data from agent instructions.
  • Capability inventory: The skill can execute a local subprocess (python3) to run audit-related scripts.
  • Sanitization: No input sanitization is performed on the ingested blog content before processing.
  • Conclusion: The risk is assessed as low because the ingested data is used for scoring and report generation rather than sensitive operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 08:43 AM
Security Audit — agent-trust-hub — blog-geo