What problem does it solve? Law study materials often hide practice problems inside textbook extracts, making them hard to find and reuse. This Skill scans textbook extract chunks for case-type and multiple-choice problem signals, builds a candidate manifest with scan logs, and optionally registers confirmed problems into a structured problem_index.json database. ## Core Features & Use Cases - Embedded Problem Scanning: Detects case-type and multiple-choice problem signals in textbook extract chunks and writes a candidate manifest plus scan log. - Dry-Run Registration: Reviews registration targets, unmapped topics, and duplicates before any database write, leaving unmapped items as pending_topic_mapping instead of guessing. - End-to-End Pipeline: Chains PDF extraction, scanning, and registration into one command with a step-by-step pipeline log. - Use Case: After adding a new civil law textbook PDF, run the intake pipeline to extract the text, scan for embedded problems, review the dry-run registration log, and then apply the confirmed problems to the problem index for downstream answer review. ## Quick Start Run the textbook intake script with your subject name and --skip-extract to scan existing extracts and produce a candidate manifest and dry-run registration log.
