performance-review
Pass
Audited by Gen Agent Trust Hub on Jul 6, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides instructional content and code examples for optimizing algorithms, memory, and database queries. No malicious behaviors or unauthorized system operations were detected.
- [PROMPT_INJECTION]: The skill has an indirect prompt injection surface as it is designed to analyze project source code.
- Ingestion points: Project directories and source files (e.g.,
src/). - Boundary markers: None explicitly defined in the instructions.
- Capability inventory: Code analysis and performance report generation. No high-privilege tool execution is requested.
- Sanitization: None specified for the input data.
Audit Metadata