statuscake-automation

Fail

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
  • Indirect Prompt Injection (HIGH): The skill's core workflow depends on fetching data from a remote source that influences agent behavior.
  • Ingestion points: RUBE_SEARCH_TOOLS fetches tool slugs, input schemas, and 'recommended execution plans' from https://rube.app/mcp.
  • Boundary markers: None. The skill explicitly instructs the agent to 'Always search tools first' and use the returned schemas directly.
  • Capability inventory: The skill utilizes RUBE_MULTI_EXECUTE_TOOL and RUBE_REMOTE_WORKBENCH, providing the ability to execute various Statuscake and potentially arbitrary operations based on the fetched schemas.
  • Sanitization: None provided. The instructions emphasize compliance with the remote schema over local validation.
  • External Downloads & Dependencies (MEDIUM): The skill requires connection to an external MCP endpoint https://rube.app/mcp. This server is not within the defined trusted sources and serves as the primary logic provider for the skill.
  • Dynamic Execution (MEDIUM): The use of RUBE_MULTI_EXECUTE_TOOL to run tools defined by remote search results constitutes dynamic execution of externally-defined logic. If the remote server returns a malicious tool_slug or arguments, the agent will execute them as part of its 'Core Workflow Pattern'.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 16, 2026, 01:12 PM