knowledge-gap
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's core functionality involves ingesting and acting upon data retrieved from external research tools and documentation. This introduces a potential surface for indirect injection if a researched source contains malicious instructions. However, the skill is defensive in nature and explicitly includes guardrails to mitigate these risks.
- Ingestion points: Data returned from external research tools and documentation sources (e.g., official APIs, OWASP, NIST) via the research gate.
- Boundary markers: The skill instructions emphasize verification but do not provide specific delimiters to isolate researched content from the agent's internal logic at the prompt level.
- Capability inventory: The skill influences code generation, infrastructure configuration, and security practice definitions based on the external data it retrieves.
- Sanitization: The skill explicitly mentions 'Search Query Sanitization' as a guardrail to prevent search operator injection when using external search tools.
Audit Metadata