sdd-research
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a robust security model for research tasks. It explicitly requires runtime capability declarations (e.g.,
gentle-ai.sdd-research-capability/v1) and forbids the agent from assuming capabilities based on available tools like Bash or generic MCP. This adheres to the principle of least privilege. - [SAFE]: Execution steps include mandatory verification of runtime grants and validation of claims against sources before persisting data. The 'Hard Rules' section specifically instructs the agent to keep evidence claims separate from product choices, which serves as a prompt-level boundary marker against data poisoning.
Audit Metadata