sentry-automation

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

• Third-party content exposure detected (medium risk: 0.65). The required runtime workflow uses Sentry MCP tools (e.g., issue/event/project detail retrieval), whose responses may include outsider-authored free text such as stack traces, issue titles/descriptions, and event payloads, which would be inserted into the agent’s LLM context via the tool outputs; this is not excluded by the “structured data only” carve-out.

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 and explicitly instructs calling RUBE_SEARCH_TOOLS to fetch current tool schemas at runtime, meaning content from that URL can control agent tool schemas/prompts and is a required external dependency.