langfuse-model-upsert

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

• Third-party content exposure detected (high risk: 0.70). The script calls the Langfuse public API (e.g., getExistingModel fetches ${LANGFUSE_HOST}/api/public/models?page=... and createModel reads the API response), directly ingesting model definitions (modelName, matchPattern, prices) from a third-party/public service which could contain untrusted/user-generated content used in the tool's workflow.