codemagic-builds

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

• Third-party content exposure detected (high risk: 0.80). The SKILL explicitly instructs the agent to download the Codemagic OpenAPI JSON from https://codemagic.io/api/v3/schema (via scripts/codemagic_fetch_openapi.py) and to fetch and summarize build metadata and step logs with scripts/codemagic.py (e.g., "builds get --step-logs", "builds failures"), which are untrusted/user-generated content the agent reads and that can materially influence actions like starting or canceling builds.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill explicitly instructs fetching the live OpenAPI schema from the Codemagic Swagger UI (https://codemagic.io/api/v3/schema) at runtime, and that downloaded JSON is used to determine available endpoints and drive the agent's request/behavior, so the external content directly controls the agent's instructions.