youtube-automation

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs runtime calls to public YouTube endpoints (e.g., YOUTUBE_SEARCH_YOU_TUBE and YOUTUBE_LIST_COMMENT_THREADS in the "Search YouTube Content" and "Manage Subscriptions and Comments" workflows) which fetch user-generated, untrusted video snippets, descriptions, and comments that the agent must read and use to drive subsequent API calls and actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly requires contacting https://rube.app/mcp at runtime to register/query Rube MCP tools (via RUBE_SEARCH_TOOLS) and fetch current tool schemas which directly determine the agent's available tools and runtime instructions, so this external endpoint can control agent behavior.