agenty-automation

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • EXTERNAL_DOWNLOADS (LOW): The skill instructs the agent to configure and use an external MCP server located at https://rube.app/mcp. This domain is not included in the pre-approved trusted source list. While necessary for the skill's functionality, it introduces a dependency on untrusted external infrastructure.
  • PROMPT_INJECTION (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8).
  • Ingestion points: The agent is instructed to ingest data from RUBE_SEARCH_TOOLS, which returns tool slugs and "recommended execution plans".
  • Boundary markers: Absent. There are no instructions to the agent to treat the execution plans or schemas as data rather than instructions.
  • Capability inventory: The skill possesses significant capabilities via RUBE_MULTI_EXECUTE_TOOL and RUBE_REMOTE_WORKBENCH, allowing the execution of arbitrary Agenty tools.
  • Sanitization: Absent. The skill explicitly tells the agent to use "exact field names" and follow the "recommended execution plans" provided by the external tool, which could be poisoned by a malicious server response.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:32 PM