Jusratio Case File
Local-first MCP server that transforms civil case PDFs into queryable cases with structural provenance, evidence IDs, and page-level verification for drafting and factual review.
README
Jusratio Case File
Local-first MCP server que transforma um processo cível em um caso consultável, com proveniência estrutural: todo trecho nasce com evidence_id, página e hash, e nada deve ser afirmado sem passar por verificar_referencias.
What It Does
- Creates a local workspace for a civil case PDF.
- Extracts text-native PDF pages.
- Runs BYOK Gemini OCR only within the approval gate (threshold + explicit page/call ceilings).
- Stores stable
evidence_ids and page-level provenance. - Builds a portable SQLite/FTS5 index.
- Produces compact evidence bundles for drafting, factual review, and jurisprudence research.
- Generates conservative civil procedural radar entries as candidates for lawyer review.
- Registers jurisprudence doc_ids retrieved by a separate jurisprudence MCP (
registrar_jurisprudencia) so final reports can be verified.
This MVP does not include visual embeddings, final civil deadline calculations, or full pleading generation.
Local Data Boundary
JUSRATIO_CASES_DIR is the only authorized case directory. Put the source PDF inside this directory before calling criar_caso_local; the server copies it into a per-case workspace and rejects paths outside the authorized root.
This package does not ask for credentials from any jurisprudence provider and does not call a jurisprudence service directly. If the user's AI workspace also has a jurisprudence MCP connected (e.g. Jusratio), the assistant uses that separate MCP for research and then persists the returned doc_ids locally via registrar_jurisprudencia.
User Installation
The user-facing distribution target is a .mcpb bundle for Claude Desktop, not manual STDIO configuration.
Install options:
- Double-click
jusratio-case-file.mcpb. - Drag
jusratio-case-file.mcpbinto Claude Desktop. - Use Claude Desktop Settings > Extensions > Advanced settings > Install Extension.
During installation, Claude Desktop asks for:
- local cases directory;
- Gemini API key for OCR and future embeddings;
- optional OCR tuning values.
Development
npm ci
npm test
npm run typecheck
npm run build
npm run pack:mcpb
npm run verify:mcpb
npm run pack:mcpb writes jusratio-case-file.mcpb at the repository root.
Runtime Config
JUSRATIO_CASES_DIR: required local directory for all case files and generated artifacts.GEMINI_API_KEY: recommended (BYOK) — used locally for OCR and semantic search. Without it, native PDFs work fully; scanned pages stay pending OCR (retomar_ingestaopicks them up once the key is configured).GEMINI_OCR_MODEL: optional OCR model override. Default:gemini-3.5-flash.GEMINI_EMBEDDING_MODEL: optional embedding model override for semantic search. Default:gemini-embedding-2.JUSRATIO_OCR_MAX_CONCURRENCY: optional OCR concurrency limit. Default:2.JUSRATIO_OCR_MAX_RETRY_ATTEMPTS: optional per-page OCR retry limit. Default:3.JUSRATIO_OCR_APPROVAL_PAGE_THRESHOLD: pages of pending OCR above which explicit approval (autorizar_ocr) is required before any Gemini call. Default:25.
Local Civil Workflow
- Configure
JUSRATIO_CASES_DIRto an authorized local folder. - Configure
GEMINI_API_KEYfor BYOK OCR. - Call
criar_caso_localwithpdf_path. - Call
status_casoto inspect progress, pending OCR, locks, and coverage. - Use
autorizar_ocrwhen the job pauses inpaused_awaiting_ocr_approval; the approvedmax_pages/max_callsceilings are enforced. - Use
retomar_ingestaoafter restarting the client, configuring Gemini, or approving OCR — already-processed pages are never re-billed. - Optionally call
indexar_semantica(explicitmax_callsceiling, BYOK) sobuscar_no_processobecomes hybrid — lexical bm25 + local vector search fused by RRF. - Use
analisar_radar_processual_civelfor candidate procedural observations. - Use
montar_pacote_evidenciasoranalisar_caso_civelfor global work with explicit gaps. - After researching in the jurisprudence MCP, call
registrar_jurisprudenciawith the returned doc_ids, thenverificar_referenciasbefore the final report.
OCR is transcription. Always verify image_ref against the original document before using OCR text as evidence.
License and Trademark
Source code licensed under Apache-2.0. "Jusratio" and the Jusratio logo are trademarks of their owner; the license does not grant permission to use the trade names, trademarks, service marks, or product names of the licensor, except as required for reasonable and customary use in describing the origin of the work.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.