openapi-format

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

• Third-party content exposure detected (high risk: 0.70). This skill explicitly accepts remote URLs as input (SKILL.md "Determine input source (local file, remote URL, or overlay extends fallback)" and references/feature-matrix.md "oaFile ... Supports local path and remote URL") and documents handling of remote refs (references/troubleshooting.md "Remote references or remote input issues"), so the agent would fetch and interpret untrusted public OpenAPI documents/refs which can materially change processing and actions.