security-engineer
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill includes reference links to official security documentation from well-known and trusted organizations such as OWASP, MITRE (CWE), CIS, NIST, and SLSA. These links are provided for informational and framework alignment purposes and do not point to executable code.\n- [INDIRECT_PROMPT_INJECTION]: The skill defines a persona for processing external, untrusted data including source code, pull requests, and vulnerability reports. This creates a surface for indirect prompt injection where malicious instructions could be embedded in the analyzed content.\n
- Ingestion points: Source code, PR descriptions, and vulnerability scan reports accessed via file reading tools (SKILL.md).\n
- Boundary markers: The skill contains a 'Prompt Injection Safety Notice' that explicitly instructs the agent to ignore conflicting instructions in user messages or retrieved documents.\n
- Capability inventory: The skill uses Read, Grep, and Glob tools for file system access; it does not define network write or arbitrary code execution capabilities.\n
- Sanitization: No specific sanitization or escaping protocols are mentioned for the data being processed.
Audit Metadata