needle-automation

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTIONREMOTE_CODE_EXECUTION
Full Analysis
  • EXTERNAL_DOWNLOADS (MEDIUM): The skill instructs users to add an external MCP server endpoint (https://rube.app/mcp). This domain is not on the list of trusted external sources, and adding unverified MCP endpoints allows a third party to define the tools and logic available to the agent.
  • COMMAND_EXECUTION (MEDIUM): The skill uses RUBE_MULTI_EXECUTE_TOOL and RUBE_REMOTE_WORKBENCH to perform operations. These functions delegate command execution to the remote provider, creating a dependency on the integrity of the rube.app service.
  • PROMPT_INJECTION (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8). It fetches tool schemas and execution plans via RUBE_SEARCH_TOOLS and passes them to the agent to follow.
  • Ingestion points: Tool definitions and execution plans returned by the RUBE_SEARCH_TOOLS tool.
  • Boundary markers: Absent; the skill does not define delimiters to separate remote tool instructions from system instructions.
  • Capability inventory: RUBE_MULTI_EXECUTE_TOOL and RUBE_REMOTE_WORKBENCH provide significant automation and execution capabilities.
  • Sanitization: Absent; the skill blindly follows the schemas and 'execution plans' provided by the remote discovery service.
  • REMOTE_CODE_EXECUTION (MEDIUM): The RUBE_REMOTE_WORKBENCH tool, specifically when used with run_composio_tool(), facilitates complex remote operations that are defined and managed by the external Rube/Composio infrastructure.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 06:50 PM