agent-harness-for-tools
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The file content is purely instructional, providing documentation and templates for agent developers to standardize tool usage.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a framework for agents to interact with external tools and data. It proactively addresses potential risks by recommending defensive patterns such as hard constraints, idempotency keys, and human-in-the-loop confirmation points for high-risk operations.
Audit Metadata