runway-dev-model-routers
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches latest configuration schemas and generation guidelines from the official Runway documentation at
docs.dev.runwayml.com. - [COMMAND_EXECUTION]: Utilizes a suite of MCP tools (
list_model_routers,create_model_router,delete_model_router, etc.) to perform authenticated state changes on the user's Runway account. - [INDIRECT_PROMPT_INJECTION]: The skill processes structured data from MCP tool outputs to inform its decisions.
- Ingestion points: Data returned from
get_model_router,list_model_routers, andget_task_routingis used to describe state and reason about routing choices. - Boundary markers: None explicitly defined; relies on the agent's internal reasoning to separate tool data from instructions.
- Capability inventory: The skill has the ability to create, update, and delete model routers, and initiate generation tasks.
- Sanitization: No explicit validation of tool output is performed within the instructions.
Audit Metadata