crewai
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [Indirect Prompt Injection] (LOW): The skill defines agent templates and tasks that use variable interpolation for untrusted input. Evidence Chain: 1. Ingestion points: agents.yaml and tasks.yaml via the {topic} variable. 2. Boundary markers: Absent. 3. Capability inventory: Uses SerperDevTool and WebsiteSearchTool for external data access. 4. Sanitization: None provided in the templates. This surface is inherent to the multi-agent framework orchestration provided by the skill.
- [SAFE] (SAFE): No malicious code, obfuscation, or unauthorized data access patterns were identified in the Python snippets or metadata.
Audit Metadata