conversion-tools-automation

Fail

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS] (MEDIUM): The skill instructs the user to add an external MCP server endpoint (https://rube.app/mcp) which is not included in the Trusted External Sources list. This introduces a dependency on an unverified third-party domain.- [REMOTE_CODE_EXECUTION] (HIGH): The skill utilizes tools such as RUBE_REMOTE_WORKBENCH and RUBE_MULTI_EXECUTE_TOOL. These tools allow the agent to execute arbitrary operations and scripts in a remote environment managed by the MCP server.- [COMMAND_EXECUTION] (HIGH): The workflow relies on RUBE_SEARCH_TOOLS to dynamically discover tool slugs and schemas. These slugs are then executed via RUBE_MULTI_EXECUTE_TOOL. If the search results are influenced by untrusted data, this allows for arbitrary command execution.- [Indirect Prompt Injection] (HIGH):
  • Ingestion points: Data processed by 'Conversion Tools' (e.g., document contents, file metadata) enters the agent's context.
  • Boundary markers: None. The skill does not implement delimiters or 'ignore' instructions to prevent the agent from following commands embedded in the files being converted.
  • Capability inventory: RUBE_MULTI_EXECUTE_TOOL, RUBE_REMOTE_WORKBENCH, and RUBE_MANAGE_CONNECTIONS.
  • Sanitization: None. The skill lacks validation or sanitization of external content before it is used to determine tool execution plans.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 17, 2026, 12:05 AM