Complete document review, extraction, comparison, and reporting for large document bundles. Use when Codex must fully read and process PDFs, PPT/PPTX, XLS/XLSX, DOC/DOCX, scanned images, email attachments, archives, or mixed research and compliance bundles; when users ask for full-text checking, complete comparison, image-aware reading, OCR coverage, formal reports, or emphasize "read everything", "do not miss images", "全文核对", "完整阅读", "一字不差", or "不能漏图". This skill is optimized for large tasks that exceed context and require external memory in the current work directory.
2026-03-19