docs/getting-started.md

# Getting started

## Requirements
- Python 3.10+
- Internet connection

## Install
```bash

git clone https://github.com/mcdaqc/hugging-research

cd hugging-research

python -m venv venv

# Windows

venv\Scripts\activate

# macOS/Linux

# source venv/bin/activate

pip install -r requirements.txt

```

## Configure
Create `.env` (token is only for the inference model; tools are anonymous):
```ini

HF_TOKEN=hf_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

MODEL_ID=Qwen/Qwen3-Coder-480B-A35B-Instruct

```

## Run
```bash

python app.py

# open http://localhost:7860

```

## Notes for Spaces
- Set `HF_TOKEN` as a Space Secret if you want to choose a different inference model via `MODEL_ID`.
- Tools never use tokens; private/gated items will be marked as not accessible.