graphcms

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

• Third-party content exposure detected (high risk: 0.70). The skill's SKILL.md instructs the agent to run Membrane actions and to proxy arbitrary GraphCMS API requests (e.g., "membrane action run ..." and "membrane request CONNECTION_ID /path/to/endpoint"), which fetches potentially user-generated/untrusted content from a third-party CMS that the agent is expected to read and could materially influence subsequent actions.