Open-source legal case management with AI-assisted PDF analysis.
Source code · Issues · Contributing · MIT license
LexCore AI brings clients, legal cases, documents, and reports into one application. Built with Rust, Tuono, React, and PostgreSQL, it connects to Ollama to answer questions about uploaded PDFs.
The project is a development prototype. Its interface and domain identifiers use Portuguese; this documentation uses English.
The canonical repository is felpzw/LexCore-AI. The project was previously called Jurídico IA, with the repository name Projeto-Banco-de-Dados. Development continues here as LexCore AI.
Contributions branch from and target develop; master is reserved for releases. See repository identity and GitHub About for canonical links, attribution, and repository description and topic values.
- Client management: create, view, update, and delete individual and corporate client records.
- Case management: associate cases with clients, lawyers, statuses, courts, and categories.
- Document management: upload, view, download, update, and delete documents linked to cases.
- AI-assisted PDF analysis: select an available Ollama model and ask questions using text extracted from a stored PDF.
- Reports: visualize documents by client and case, cases by lawyer and status, and hearings by client and lawyer.
- Development utilities: initialize the database schema, load sample records, and reset a disposable database.
| Layer | Implementation |
|---|---|
| Application framework | Tuono 0.19.7 |
| Backend | Rust, Tokio, tokio-postgres, Reqwest |
| Frontend | React 19, TypeScript, Recharts |
| Database | PostgreSQL 16 |
| AI inference | Ollama |
| Local services | Docker Compose |
Clone the canonical repository and switch to the development branch:
git clone https://github.com/felpzw/LexCore-AI.git
cd LexCore-AI
git switch developFollow the setup and environment guide for prerequisites, configuration, model installation, and database initialization.
After completing the prerequisites and configuring var.env, run from the repository root:
npm ci
docker compose up -d
tuono devOpen http://localhost:3000. On a fresh database, visit /configuracoes to initialize the schema and load sample records before using the management pages.
The database reset action drops every table in the connected database's public schema. Use it only with a disposable development database.
| Guide | Contents |
|---|---|
| Documentation index | Reading order and documentation maintenance |
| Repository identity and About | Canonical links, project history, GitHub metadata, and attribution |
| Project overview | Architecture, modules, data flow, and current limitations |
| Setup and environment | Local installation, configuration, and troubleshooting |
| Git workflow | Branching, pull requests, releases, and hotfixes |
| Commit standards | Commit format, types, scopes, and examples |
docs/ Project and contributor documentation
src/app.rs Application startup and environment loading
src/lib.rs Shared database and query helpers
src/routes/ React pages and Rust route handlers
src/routes/api/ Backend API endpoints
src/components/ Reusable React components
src/styles/ Application styles
docker-compose.yml PostgreSQL and Ollama services
var.env Local runtime configuration (currently tracked)
Read the contributing guide, Git workflow, and commit standards. Open regular pull requests against develop. Keep changes focused, update the relevant documentation, and describe the checks actually performed in each pull request.
LexCore AI is licensed under the MIT License.
The application builds on open-source technologies including Tuono, React, PostgreSQL, Ollama, and Recharts. Dependencies and downloaded models retain their own licenses; see the attribution notes.