workflow
Pass
Audited by Gen Agent Trust Hub on Aug 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acts as an orchestrator for goals that may originate from untrusted sources. 1. Ingestion points: Processes user-supplied goals through triggers like '/workflow' or 'orchestrate X'. 2. Boundary markers: Relies on external SOPs and guardrails like CLAUDE.md rather than explicit data delimiters. 3. Capability inventory: The workflow engine can invoke other skills, execute tool scripts, or make MCP calls. 4. Sanitization: Employs an autonomy gate to require approval for high-risk actions and a draft-never-send policy for sensitive external communication.
Audit Metadata