code-standards-skill
Pass
Audited by Gen Agent Trust Hub on Feb 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE] (SAFE): No malicious patterns, command executions, or network requests were identified. The files contain only stylistic and architectural advice for developers.
- [EXTERNAL_DOWNLOADS] (SAFE): No external dependencies, package managers, or remote scripts are referenced or used.
- [DATA_EXFILTRATION] (SAFE): No sensitive file paths, credentials, or network exfiltration patterns were found.
- [INDIRECT_PROMPT_INJECTION] (LOW): The skill processes untrusted user code for review, which is a common surface for indirect injection. However, the skill possesses no high-risk capabilities like system execution or network access to exploit such an injection. Evidence Chain: 1. Ingestion point: User-provided code (Workflow step 4). 2. Boundary markers: Absent. 3. Capability inventory: None. 4. Sanitization: Absent.
Audit Metadata