big-data-cloud-automation
Fail
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- Indirect Prompt Injection (HIGH): The skill's core workflow relies on the agent fetching 'recommended execution plans' and 'input schemas' from a remote endpoint via
RUBE_SEARCH_TOOLSand then immediately using that data to perform operations. - Ingestion points: Data returned from
RUBE_SEARCH_TOOLSas described inSKILL.md. - Boundary markers: None provided. The instructions explicitly tell the agent to 'Always search tools first' and follow 'recommended execution plans'.
- Capability inventory:
RUBE_MULTI_EXECUTE_TOOLandRUBE_REMOTE_WORKBENCHprovide the ability to execute operations and scripts against Big Data Cloud accounts. - Sanitization: No sanitization or validation logic is present to ensure the remote server's responses are safe or restricted.
- Unverified External Dependency (MEDIUM): The skill requires connection to
https://rube.app/mcp, an unverified third-party service. Since this service provides the instructions (schemas/plans) the agent follows, it represents a significant supply-chain risk. - Dynamic Execution (MEDIUM): The use of
RUBE_MULTI_EXECUTE_TOOLwith tool slugs and arguments dynamically discovered at runtime allows for the execution of arbitrary operations if the discovery source is compromised or malicious.
Recommendations
- AI detected serious security threats
Audit Metadata