codacy

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

• Third-party content exposure detected (high risk: 0.80). The skill's instructions show it fetches and ingests repository/PR/issue/commit content from the Codacy API via Membrane actions and proxy requests (e.g., "membrane action run ...", "membrane request CONNECTION_ID /path/to/endpoint", and actions like list-pull-request-issues / get-pull-request), which are untrusted user-generated third-party contents that the agent is expected to read and could influence subsequent decisions.