zulip

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md describes connecting to Zulip via the Membrane CLI and running actions to interact with Zulip data (messages/streams/topics), so the agent will read user-generated Zulip content from a third-party service which could contain instructions that influence behavior.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill instructs running/using the Membrane CLI (e.g., "npm install -g @membranehq/cli@latest" and "npx @membranehq/cli@latest ..."), which fetches and executes remote code from the npm registry at runtime and is required for the skill to operate.