percentile-analyzer
Pass
Audited by Gen Agent Trust Hub on Feb 18, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE] (SAFE): The analyzed skill contains only a markdown definition with metadata and no executable scripts or code blocks.
- [COMMAND_EXECUTION] (SAFE): The skill allows the 'Bash' tool, representing a high-privilege capability; however, no actual commands are present to evaluate for malicious intent.
- [INDIRECT_PROMPT_INJECTION] (LOW): The skill is designed to process external performance data, creating a potential surface for indirect injection. 1. Ingestion points: Performance logs or test results (SKILL.md). 2. Boundary markers: Absent from instructions. 3. Capability inventory: Bash, Read, Write, Edit. 4. Sanitization: No sanitization logic provided.
Audit Metadata