trello-automation

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

• Third-party content exposure detected (high risk: 1.00). The SKILL.md explicitly instructs the agent to call Trello endpoints (e.g., TRELLO_GET_SEARCH, TRELLO_GET_MEMBERS_BOARDS_BY_ID_MEMBER, TRELLO_GET_BOARDS_BY_ID_BOARD, and endpoints that fetch card descriptions, comments, and attachments) to read user-generated board/card content as part of workflows like "Search and Filter Cards" and "Add Comments and Attachments", so untrusted third‑party content is ingested and can influence subsequent tool 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 connecting to an external MCP server at https://rube.app/mcp and calling RUBE_SEARCH_TOOLS/RUBE_MANAGE_CONNECTIONS at runtime to fetch tool schemas that determine available tools and agent behavior, so this external URL can directly control prompts/instructions.