dagster-development

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

• Third-party content exposure detected (high risk: 0.70). The skill includes runtime examples that ingest untrusted third-party content — e.g., NASAResource making requests.get calls to https://api.nasa.gov and the adhoc_request_sensor that json.load()s .json files from a data/requests directory and uses them to build run_config — so the agent would read and act on external/public or user-provided data.