image-parser

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

• Third-party content exposure detected (high risk: 0.70). This skill sends user images to the third‑party SoMark API (SOMARK_SYNC_URL https://somark.tech/api/v1/parse/sync) and ingests the returned OCR/text outputs (e.g., *.text_bbox.json) — untrusted/user-generated content the agent is expected to read and that can influence downstream extraction or automation.