respondio

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs using Membrane's proxy to call the Respond.io API (e.g., "membrane request CONNECTION_ID /path/to/endpoint"), which fetches untrusted, user-generated conversation/messages from a public third‑party service that the agent is expected to read and act on, enabling indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill requires installing/running the Membrane CLI via npm/npx (e.g., npm install -g @membranehq/cli / npx @membranehq/cli@latest ...), which at runtime fetches and executes remote package code from the npm registry (e.g. https://registry.npmjs.org/@membranehq/cli), so it is a runtime external dependency that executes remote code.