dingconnect

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

• Third-party content exposure detected (high risk: 0.80). The skill instructs the agent to use the Membrane CLI to run actions against a DingConnect connection (e.g., membrane action list / membrane action run), which returns data from DingConnect entities such as Knowledge Base Articles, messages, templates and other user-generated content that the agent is expected to read and which can materially influence subsequent actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill requires and runs the Membrane CLI fetched at runtime (e.g., "npx @membranehq/cli@latest" / "npm install -g @membranehq/cli@latest"), which downloads and executes remote code and is relied on to drive the agent's actions.