crewai-agent-workflow
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions and associated reference materials do not contain any malicious patterns or unsafe commands. The implementation templates promote secure agent design.
- [PROMPT_INJECTION]: The skill requires implementers to maintain a prompt-injection posture across all agent and task messages, aligning with security standards to prevent instruction override.
- [INDIRECT_PROMPT_INJECTION]: The skill acknowledges that workflows ingest external data (e.g., source material in the Research-and-synthesize reference) and addresses this surface by mandating code-enforced termination authority and evaluation gates for grounding and citation accuracy.
- [REMOTE_CODE_EXECUTION]: No remote code execution patterns were found. The skill references the well-known CrewAI framework for orchestration.
Audit Metadata