design-task
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of educational documentation and configuration examples for designing tasks within the CrewAI framework.
- [INDIRECT_PROMPT_INJECTION]: The documentation describes patterns for ingesting external data via tools (e.g., ScrapeWebsiteTool) and variable interpolation. While this presents an attack surface for indirect prompt injection, the skill proactively recommends mitigation strategies such as task guardrails, specific expected_output definitions, and human-in-the-loop review (human_input=True).
Audit Metadata