release-management
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- PROMPT_INJECTION (SAFE): Minimal risk of indirect prompt injection from repository data.
- Ingestion points: Processes git logs and external documentation mentioned in SKILL.md.
- Boundary markers: No explicit delimiters are defined for untrusted project data.
- Capability inventory: Authorized to use bash and git tools for release tasks.
- Sanitization: No custom sanitization logic is implemented, relying on the agent's core safety protocols.
Audit Metadata