| title: BuildTrack Local | |
| emoji: 🏗️ | |
| colorFrom: blue | |
| colorTo: green | |
| sdk: docker | |
| app_port: 7860 | |
| pinned: false | |
| # BuildTrack Local | |
| A fully local construction project management system. The assistant, document suggestions, BOQ parsing, risk scoring, and project insights run with deterministic local logic inside the app. No Gemini, OpenAI, Anthropic, Hugging Face Inference API, or other hosted AI key is required. | |
| ## Run Locally | |
| **Prerequisites:** Node.js 20+ | |
| 1. Install dependencies: | |
| `npm install` | |
| 2. Optional: copy [.env.example](.env.example) to `.env.local` and set `MONGODB_URI` or `JWT_SECRET`. | |
| 3. Start the app: | |
| `npm run dev` | |
| 4. Open `http://localhost:7860` unless you set a different `PORT`. | |
| If MongoDB is not available, the server automatically uses an in-memory local database for the current session. | |
| ## Deploy To Hugging Face Spaces | |
| Create a new Hugging Face Space using the **Docker** SDK, then upload this repository. The included [Dockerfile](Dockerfile) builds the Vite frontend and starts the local Express server on the Space port. | |
| No AI environment variables are needed. Optional variables: | |
| - `MONGODB_URI` - connect to your own MongoDB instance. Leave blank for in-memory storage. | |
| - `JWT_SECRET` - custom signing secret for login tokens. | |
| - `PORT` - Hugging Face sets this automatically. | |