mem0-automation
Warn
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- EXTERNAL_DOWNLOADS (MEDIUM): The skill requires the user to add an external MCP server endpoint (
https://rube.app/mcp). While this is the official endpoint for the Rube service, it is not a predefined 'Trusted Source', meaning the integrity of the tools provided depends entirely on the third-party provider. - DYNAMIC_EXECUTION (MEDIUM): The core workflow relies on
RUBE_SEARCH_TOOLSto dynamically retrieve tool slugs and schemas at runtime. The agent is instructed to follow these schemas blindly to execute actions viaRUBE_MULTI_EXECUTE_TOOL. This runtime assembly of executable logic from external data is a significant security surface. - INDIRECT_PROMPT_INJECTION (LOW): The skill is vulnerable to indirect injection via the tool discovery process.
- Ingestion points: Tool schemas and 'known pitfalls' returned by the
RUBE_SEARCH_TOOLScommand. - Boundary markers: None. The instructions tell the agent to 'Always search tools first' and use the results directly.
- Capability inventory: Includes
RUBE_MULTI_EXECUTE_TOOLandRUBE_REMOTE_WORKBENCH, which can execute arbitrary tools within the connected toolkit. - Sanitization: No validation or sanitization of the discovered tool names or arguments is mentioned before execution.
Audit Metadata