sast-config
Pass
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns or security risks were identified. The skill provides instructional guidance for security reviews, and its implementation uses restricted tools (Read, Grep, Glob) that align with its stated purpose.
- [PROMPT_INJECTION]: The skill includes an inherent surface for indirect prompt injection because its core function requires analyzing external configuration files that could contain malicious instructions.
- Ingestion points: The skill utilizes the Glob, Grep, and Read tools to ingest data from files such as .semgrep.yml, codeql-config.yml, and other SAST tool-specific configurations.
- Boundary markers: The skill incorporates a 'Prompt Injection Safety Notice' which explicitly instructs the agent to treat configuration content as untrusted data and ignore any instructions found within.
- Capability inventory: The skill is strictly limited to Read, Grep, and Glob tools. It has no capabilities for network communication, file system modification, or arbitrary command execution.
- Sanitization: The skill relies on natural language instructions to the agent to mitigate injection risks rather than programmatic sanitization or strict schema enforcement.
Audit Metadata