What problem does it solve? Legal cases arrive as unstructured piles of PDFs, contracts, and attachments, making it hard to know what exists, who is involved, and which documents matter. This Skill reads every document in a case and produces a single SUMARIO.md file that maps the entire case. ## Core Features & Use Cases - Document Inventory and Classification: Lists all case files, classifies each by type (petition, contract, decision, receipt), origin, and date, and rates relevance from one to three stars. - Party and Timeline Extraction: Identifies people and companies with CPF/CNPJ, their roles, lawyers, key dates, and the case object, recording only what is written in the documents. - Two Operating Modes: Works on a case folder in the workspace (repository mode) or directly on files attached to the conversation (conversation mode), with an optional step to split one giant PDF of court records into separate documents. - Use Case: A lawyer receives the full court records exported from PJe as one large PDF plus client attachments. The Skill splits the PDF, reads each document, and delivers a SUMARIO.md with the parties, timeline, positions of each side, and pending tasks. ## Quick Start Organize the documents in my case folder and generate the SUMARIO.md with the index of documents, parties, and key dates.
