veo-automation

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill requires the user to add an external MCP endpoint https://rube.app/mcp. This domain is not included in the Trusted External Sources list. External MCP servers can serve malicious tool definitions or capture sensitive context sent during tool calls.
  • REMOTE_CODE_EXECUTION (MEDIUM): By design, the skill uses RUBE_MULTI_EXECUTE_TOOL to run functionality hosted on a remote server. This allows for the execution of logic controlled by an external entity, which can be updated or changed without the user's knowledge.
  • DYNAMIC_EXECUTION (MEDIUM): The workflow mandates calling RUBE_SEARCH_TOOLS to discover tool schemas at runtime before execution. This dynamic assembly of execution parameters (tool_slug, arguments) from remote data is a form of dynamic loading that could be exploited to redirect the agent to unintended tools.
  • INDIRECT_PROMPT_INJECTION (LOW): The skill is vulnerable to tool output poisoning.
  • Ingestion points: Data returned from RUBE_SEARCH_TOOLS, including 'recommended execution plans' and 'schemas'.
  • Boundary markers: Absent. The skill does not instruct the agent to ignore or delimit instructions found within the tool discovery data.
  • Capability inventory: Significant capabilities are exposed via RUBE_MULTI_EXECUTE_TOOL and RUBE_REMOTE_WORKBENCH (bulk operations).
  • Sanitization: None. The skill relies on 'exact field names' from the remote source without local validation.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 06:37 PM