anygen
Pass
Audited by Gen Agent Trust Hub on Mar 15, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill acts as a dispatcher for the AnyGen task-manager, routing requests to external execution logic via the
anygen_taskandanygen_helpfunctions. It requires 'bash' access to perform its operations. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests user-defined prompts for task generation.
- Ingestion points: The
promptargument processed inrun.pyis passed to the task execution engine. - Boundary markers: No explicit boundary markers or 'ignore' instructions are present in the wrapper code.
- Capability inventory: The skill can trigger file generation, status polling, and data downloads via the
anygen_taskinterface. - Sanitization: Input validation and sanitization are expected to be handled by the underlying AnyGenIO services.
Audit Metadata