company-research
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to fetch and process data from external websites and user-provided documents, creating an attack surface for instructions hidden in third-party content.
- Ingestion points: SKILL.md instructs the agent to gather facts from web searches, official company pages, and documents provided by the user.
- Boundary markers: The skill uses a rigorous 'Source Tier' hierarchy (defined in references/source-tiers.md) to separate confirmed facts from unverified reports, providing a conceptual boundary for the agent.
- Capability inventory: The skill can perform web searches, read files, and execute a local utility script (scripts/check_source_coverage.py).
- Sanitization: The validation script (scripts/check_source_coverage.py) enforces strict type and schema validation on the JSON data generated by the agent before it is finalized.
Audit Metadata