outlook-automation

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

• Third-party content exposure detected (medium risk: 0.30). [Outlook automation skill workflow “Search and Filter Emails” reads mailbox email contents via OUTLOOK_SEARCH_MESSAGES (and optionally OUTLOOK_GET_MESSAGE / attachment download), so outsider-authored free text that the user/attacker has posted into the mailbox can be ingested at runtime.]

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

• Potentially malicious external URL detected (high risk: 0.90). The skill requires adding and using the MCP endpoint https://rube.app/mcp at runtime (see "Always call RUBE_SEARCH_TOOLS first to get current tool schemas"), which means remote tool schemas from that URL will be fetched and can directly control agent behavior and instructions.