validation
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of instructional markdown and templates for scientific research planning. It contains no executable scripts, system commands, or network operations.
- [INDIRECT_PROMPT_INJECTION]: The skill accepts research hypotheses as input to populate its experiment templates. 1. Ingestion points: User-provided hypotheses within the experimental design workflow. 2. Boundary markers: Absent. 3. Capability inventory: The skill has no access to shell commands, file system writes, or network requests. 4. Sanitization: None. Despite the lack of sanitization, the absence of dangerous tools ensures the ingestion surface is safe.
Audit Metadata