secrets-management
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill establishes secure patterns for credential management. * Promotes the use of .gitignore to prevent sensitive file exposure. * Uses placeholder values for API key format documentation. * Encourages environment variable fallbacks as a standard security practice.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing local configuration data. * Ingestion points: SKILL.md contains a Python pattern for reading secrets.json into the agent's context. * Boundary markers: Absent; the data is parsed as standard JSON. * Capability inventory: The provided scripts are limited to file system read access for configuration and environment variable retrieval. * Sanitization: Uses standard json.load() without specific value sanitization.
Audit Metadata