Spaces:
Sleeping
Sleeping
|
Download README.md from umar-30/Multi-agent-system: direct link, hf CLI and curl.
- Browser
- Download file 909 Bytes
-
https://huggingface.co/spaces/umar-30/Multi-agent-system/resolve/main/README.md
- Command line
-
hf download hf://spaces/umar-30/Multi-agent-system/README.md
-
curl -L -o README.md https://huggingface.co/spaces/umar-30/Multi-agent-system/resolve/main/README.md
909 Bytes
| title: AI Multi-Agent System | |
| emoji: ⚡ | |
| colorFrom: indigo | |
| colorTo: blue | |
| sdk: docker | |
| app_port: 7860 | |
| pinned: false | |
| # AI Multi-Agent System | |
| An advanced orchestration platform where specialized AI agents work together to solve complex research and coding tasks. | |
| ## Features | |
| - **Research Agent**: Deep exploration and information gathering. | |
| - **Coding Agent**: Automated Python solution generation. | |
| - **Summary Agent**: Clear and concise final explanations. | |
| - **Modern UI**: Dark-themed dashboard built with Streamlit. | |
| ## How to Run Locally | |
| 1. Clone the repository. | |
| 2. Install dependencies: `pip install -r requirements.txt` | |
| 3. Create a `.env` file and add your `COHERE_API_KEY`. | |
| 4. Run the app: `streamlit run app.py` | |
| ## Deployment | |
| This project is configured for deployment on **Hugging Face Spaces**. | |
| Ensure you set the `COHERE_API_KEY` in the Space's Secret settings. | |