grilling
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions create an attack surface for indirect prompt injection by directing the agent to ingest untrusted data from the environment.
- Ingestion points: In
SKILL.md, the instructions tell the agent to "Look up discoverable facts in the available code, documentation, and tools." - Boundary markers: The skill lacks explicit boundary markers or instructions to treat data from external files as untrusted content, increasing the risk that embedded instructions in those files could be obeyed.
- Capability inventory: The agent has the ability to delegate investigations, perform "documentation writes," and execute "external actions" if authorized in the initial request.
- Sanitization: There are no defined mechanisms for sanitizing, filtering, or escaping the content retrieved from external files before it is processed by the agent.
Audit Metadata