orchestrate-governance-loop
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is a meta-orchestrator that sequences other internal skills within the same vendor ecosystem. No evidence of malicious intent or safety bypasses was found.
- [COMMAND_EXECUTION]: The skill manages the execution flow of sub-skills and performs filesystem writes to a specific local path (
docs/calibration/) to store the governance report. This behavior is expected and restricted to the project environment. - [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection. 1. Ingestion points: Free-form project context and trigger event data in
SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: File-write todocs/calibration/and execution of sub-skills inSKILL.md. 4. Sanitization: Absent. The risk is neutralized by the skill's instructions to use the input for metadata classification only, preventing the input from directly influencing the logic beyond logging.
Audit Metadata