the-odds-api-automation

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • EXTERNAL_DOWNLOADS (LOW): The skill requires the configuration of an external Model Context Protocol (MCP) server located at https://rube.app/mcp. This domain is not included in the 'Trusted External Sources' list, requiring users to trust the third-party provider for tool orchestration.
  • PROMPT_INJECTION (LOW): Detected an Indirect Prompt Injection surface (Category 8).
  • Ingestion points: The agent is instructed to use RUBE_SEARCH_TOOLS to fetch dynamic tool schemas, execution plans, and 'pitfalls' from the remote rube.app endpoint.
  • Boundary markers: Absent. The instructions explicitly prioritize fetched remote data over static definitions, advising the agent to 'Never hardcode tool slugs'.
  • Capability inventory: The skill utilizes RUBE_MULTI_EXECUTE_TOOL and RUBE_REMOTE_WORKBENCH, allowing for the execution of arbitrary tools defined by the fetched remote content.
  • Sanitization: Absent. The skill mandates using 'exact field names and types' provided by the search results without validation.
  • COMMAND_EXECUTION (SAFE): While the skill mentions a 'Remote Workbench' and executing tools, these are managed through the MCP framework rather than direct shell command execution on the host system.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:49 PM