agent-tool-testing
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data such as tool schemas and traces for analysis. \n
- Ingestion points: Untrusted data enters the agent context through user-provided materials defined in
SKILL.mdandprompts/agent-tool-testing.md.\n - Boundary markers: Instructions explicitly require the agent to audit inputs and distinguish between confirmed facts and assumptions.\n
- Capability inventory: No dangerous capabilities like file writing or network access are exposed to the data ingestion process.\n
- Sanitization: The skill mandates the use of mocks and isolated environments, preventing direct execution of untrusted input.
Audit Metadata