workflow-orchestrator
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill orchestrates workflows from broad requests and delegates tasks to subagents, creating a multi-step execution chain.
- Ingestion points: The discovery stage processes external objectives to create the workflow (SKILL.md).
- Boundary markers: Instructions specify "output contracts" and "bounded scope" for deliverables, though they do not mandate technical isolation between data and instructions.
- Capability inventory: The skill is restricted by the "read-only" sandbox metadata and contains no executable scripts or command-line operations.
- Sanitization: The workflow includes "validation checkpoints" and "final verification" stages to check deliverables before integration.
Audit Metadata