ccf-experiment-designer
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided ideas, claims, and experimental results to generate protocols and tables. This represents an ingestion surface for untrusted data. However, the instructions provide strong integrity safeguards, specifically a 'no-fabrication' policy and requirements to preserve real values. There are no high-risk capabilities such as arbitrary network access or system-wide command execution enabled for this data processing.
- [SAFE]: The skill's primary function is to provide structure for scientific documentation and planning. It uses a restricted working directory ('ccfa-workfiles/') and adheres to established scientific best practices without employing dangerous execution patterns or obfuscation.
Audit Metadata