experiment-protocol
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill is composed entirely of markdown instructions and flowcharts designed to guide agent behavior. It does not include any scripts, binaries, or hidden executable content.
- [SAFE]: The skill does not access sensitive files, perform network operations outside the provided MCP server scope, or attempt to bypass security guardrails. All operations are restricted to the documented experiment lifecycle tools.
Audit Metadata