calibrate
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or dangerous execution methods were identified in the skill instructions or metadata. The skill focuses on organizational alignment and documentation.
- [INDIRECT_PROMPT_INJECTION]: The skill involves auditing and standardizing codebase components, which is a common ingestion surface for untrusted data.
- Ingestion points: Processes workflow components, naming conventions, and error formats from the existing codebase (SKILL.md).
- Boundary markers: None explicitly defined in the skill text.
- Capability inventory: Identifies deviations and updates documentation (.maestro.md), utilizing standard agent file-system access.
- Sanitization: Not present, as the skill is an instructional framework for the agent.
Audit Metadata