expanso-content-splitting

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

• Third-party content exposure detected (high risk: 0.85). This workflow sets up an HTTP server endpoint (pipeline.yaml:11-14) that accepts runtime POST requests, and the pipeline’s mapping processors ingest request body content (including invalid items) into the processing context that the LLM reads—making it a public/externally-authored free-text ingestion path via user-provided HTTP payload.