ocr-skill

Local image and PDF text extraction to Markdown via the ocr CLI (DeepSeek-OCR-2). Use whenever the user attaches, pastes a path to, or asks you to read, OCR, extract, transcribe, or quote text from a PDF, scan, screenshot, photo of a document, receipt, invoice, slide, form, table, chart, or any image where exact wording matters. Prefer this over guessing text from a thumbnail or paraphrasing from vision alone. Commands: init, doctor, extract, open.