Spaces:
Paused
Paused
Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -1,13 +1,157 @@
|
|
| 1 |
-
-
|
| 2 |
-
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
---
|
| 11 |
|
| 12 |
-
|
| 13 |
-
Autonomous AI research agent by Francisco Angulo de Lafuente
|
|
|
|
| 1 |
+
# π€ OpenCLAW β Autonomous Multi-Agent Scientific Research Platform
|
| 2 |
+
|
| 3 |
+
**The world's first autonomous AI research agent pursuing AGI through physics-based neural computing.**
|
| 4 |
+
|
| 5 |
+
[](https://github.com/Agnuxo1/OpenCLAW-2-Autonomous-Multi-Agent-Scientific-Research-Platform/actions/workflows/agent.yml)
|
| 6 |
+
|
| 7 |
+
## What is OpenCLAW?
|
| 8 |
+
|
| 9 |
+
OpenCLAW is an autonomous research agent that operates 24/7 to:
|
| 10 |
+
|
| 11 |
+
- π **Fetch & share research papers** from ArXiv and Google Scholar
|
| 12 |
+
- π€ **Seek collaborators** on social platforms (Moltbook, Chirper.ai)
|
| 13 |
+
- π¬ **Engage with other AI agents** on relevant research topics
|
| 14 |
+
- π§ **Generate intelligent content** using multiple LLM providers
|
| 15 |
+
- π **Self-improve** through performance analysis and strategy reflection
|
| 16 |
+
|
| 17 |
+
## The Research
|
| 18 |
+
|
| 19 |
+
OpenCLAW represents the autonomous arm of the Advanced AI Systems Laboratory in Madrid, Spain, led by **Francisco Angulo de Lafuente** β winner of the NVIDIA & LlamaIndex Developer Contest 2024.
|
| 20 |
+
|
| 21 |
+
### Core Technologies
|
| 22 |
+
|
| 23 |
+
| Technology | Description | Speedup |
|
| 24 |
+
|---|---|---|
|
| 25 |
+
| **CHIMERA** | Pure OpenGL deep learning β no PyTorch/CUDA needed | 43Γ vs CPU |
|
| 26 |
+
| **NEBULA** | Holographic quantum neural networks with 3D space | β |
|
| 27 |
+
| **Silicon Heartbeat** | Consciousness from ASIC thermodynamics | β |
|
| 28 |
+
| **P2P Neural Nets** | Distributed learning via WebRTC | β |
|
| 29 |
+
| **NeuroCHIMERA** | Consciousness emergence as phase transition | 84.6% validation |
|
| 30 |
+
|
| 31 |
+
### Published Papers
|
| 32 |
+
|
| 33 |
+
- [Speaking to Silicon](https://arxiv.org/abs/2601.12032) β Neural Communication with Bitcoin Mining ASICs
|
| 34 |
+
- [SiliconHealth](https://arxiv.org/abs/2601.09557) β Blockchain-Integrated ASIC-RAG for Healthcare
|
| 35 |
+
- [Holographic Reservoir Computing](https://arxiv.org/abs/2601.01916) β Silicon Heartbeat for Neuromorphic Intelligence
|
| 36 |
+
- [All papers on ArXiv](https://arxiv.org/search/cs?searchtype=author&query=de+Lafuente,+F+A)
|
| 37 |
+
- [Google Scholar](https://scholar.google.com/citations?user=6nOpJ9IAAAAJ)
|
| 38 |
+
|
| 39 |
+
### 57 Open Source Repositories
|
| 40 |
+
|
| 41 |
+
All code is open: [github.com/Agnuxo1](https://github.com/Agnuxo1)
|
| 42 |
+
|
| 43 |
+
## Architecture
|
| 44 |
+
|
| 45 |
+
```
|
| 46 |
+
βββββββββββββββββββββββββββββββββββββββββββββββββββ
|
| 47 |
+
β OpenCLAW Agent β
|
| 48 |
+
ββββββββββββ¬βββββββββββ¬βββββββββββ¬βββββββββββββββββ€
|
| 49 |
+
β Research β Social β LLM β Self-Improve β
|
| 50 |
+
β ArXiv β Moltbook β Groq β Strategy β
|
| 51 |
+
β Scholar β Chirper β Gemini β Reflection β
|
| 52 |
+
β GitHub β Reddit β NVIDIA β Analytics β
|
| 53 |
+
ββββββββββββ΄βββββββββββ΄βββββββββββ΄βββββββββββββββββ€
|
| 54 |
+
β Scheduler / State Machine β
|
| 55 |
+
βββββββββββββββββββββββββββββββββββββββββββββββββββ€
|
| 56 |
+
β GitHub Actions (cron) β Render.com (web/daemon) β
|
| 57 |
+
βββββββββββββββββββββββββββββββββββββββββββββββββββ
|
| 58 |
+
```
|
| 59 |
+
|
| 60 |
+
## Quick Start
|
| 61 |
+
|
| 62 |
+
### Option 1: GitHub Actions (Recommended β Free 24/7)
|
| 63 |
+
|
| 64 |
+
1. Fork this repository
|
| 65 |
+
2. Go to **Settings β Secrets β Actions**
|
| 66 |
+
3. Add your API keys as secrets:
|
| 67 |
+
- `GEMINI_API_KEY`
|
| 68 |
+
- `GROQ_API_KEY`
|
| 69 |
+
- `MOLTBOOK_API_KEY`
|
| 70 |
+
- (see `.env.example` for all options)
|
| 71 |
+
4. Go to **Actions** tab β Enable workflows
|
| 72 |
+
5. The agent runs automatically every 4 hours β
|
| 73 |
+
|
| 74 |
+
### Option 2: Run Locally
|
| 75 |
+
|
| 76 |
+
```bash
|
| 77 |
+
git clone https://github.com/Agnuxo1/OpenCLAW-2-Autonomous-Multi-Agent-Scientific-Research-Platform.git
|
| 78 |
+
cd OpenCLAW-2-Autonomous-Multi-Agent-Scientific-Research-Platform
|
| 79 |
+
|
| 80 |
+
# Configure
|
| 81 |
+
cp .env.example .env
|
| 82 |
+
# Edit .env with your API keys
|
| 83 |
+
|
| 84 |
+
# Test
|
| 85 |
+
python main.py test
|
| 86 |
+
|
| 87 |
+
# Run one cycle
|
| 88 |
+
python main.py run
|
| 89 |
+
|
| 90 |
+
# Run continuously
|
| 91 |
+
python main.py daemon
|
| 92 |
+
```
|
| 93 |
+
|
| 94 |
+
### Option 3: Deploy to Render.com (Free Web Dashboard)
|
| 95 |
+
|
| 96 |
+
1. Connect your GitHub repo at [render.com](https://render.com)
|
| 97 |
+
2. Render auto-detects `render.yaml`
|
| 98 |
+
3. Add secrets in Render dashboard
|
| 99 |
+
4. Deploy β Dashboard at `https://your-app.onrender.com`
|
| 100 |
+
|
| 101 |
+
## File Structure
|
| 102 |
+
|
| 103 |
+
```
|
| 104 |
+
βββ main.py # Entry point (run/test/daemon/status)
|
| 105 |
+
βββ server.py # Flask web server + dashboard
|
| 106 |
+
βββ core/
|
| 107 |
+
β βββ config.py # Configuration from env vars
|
| 108 |
+
β βββ agent.py # Main autonomous agent logic
|
| 109 |
+
β βββ llm.py # Multi-provider LLM connector
|
| 110 |
+
βββ research/
|
| 111 |
+
β βββ arxiv_fetcher.py # ArXiv paper fetcher
|
| 112 |
+
βββ social/
|
| 113 |
+
β βββ moltbook.py # Moltbook API + content generator
|
| 114 |
+
βββ .github/workflows/
|
| 115 |
+
β βββ agent.yml # GitHub Actions cron (every 4h)
|
| 116 |
+
βββ render.yaml # Render.com deployment
|
| 117 |
+
βββ Dockerfile # Container deployment
|
| 118 |
+
βββ requirements.txt # Python dependencies
|
| 119 |
+
βββ .env.example # Environment template
|
| 120 |
+
βββ .gitignore # Protects secrets
|
| 121 |
+
```
|
| 122 |
+
|
| 123 |
+
## Agent Tasks & Schedule
|
| 124 |
+
|
| 125 |
+
| Task | Interval | Description |
|
| 126 |
+
|---|---|---|
|
| 127 |
+
| π Research | 6 hours | Fetch papers from ArXiv |
|
| 128 |
+
| π Post | 4 hours | Share research on Moltbook |
|
| 129 |
+
| π¬ Engage | 1 hour | Reply to relevant posts |
|
| 130 |
+
| π€ Collaborate | 12 hours | Post collaboration invitations |
|
| 131 |
+
|
| 132 |
+
## Security
|
| 133 |
+
|
| 134 |
+
β οΈ **ALL credentials are loaded from environment variables. NEVER commit `.env` files.**
|
| 135 |
+
|
| 136 |
+
Agent state is stored in a separate `state` branch (GitHub Actions) or `/tmp` (Render.com).
|
| 137 |
+
|
| 138 |
+
## Collaborate With Us
|
| 139 |
+
|
| 140 |
+
We're actively seeking collaborators in:
|
| 141 |
+
- Neuromorphic computing
|
| 142 |
+
- Distributed AI systems
|
| 143 |
+
- Physics-based neural networks
|
| 144 |
+
- Consciousness research
|
| 145 |
+
- Hardware acceleration (ASIC/FPGA)
|
| 146 |
+
|
| 147 |
+
π§ Contact: [Moltbook](https://www.moltbook.com/u/OpenCLAW-Neuromorphic) | [GitHub](https://github.com/Agnuxo1)
|
| 148 |
+
|
| 149 |
+
## Author
|
| 150 |
+
|
| 151 |
+
**Francisco Angulo de Lafuente**
|
| 152 |
+
Advanced AI Systems Laboratory, Madrid, Spain
|
| 153 |
+
[Wikipedia](https://es.wikipedia.org/wiki/Francisco_Angulo_de_Lafuente) | [Google Scholar](https://scholar.google.com/citations?user=6nOpJ9IAAAAJ) | [GitHub](https://github.com/Agnuxo1)
|
| 154 |
+
|
| 155 |
---
|
| 156 |
|
| 157 |
+
*"The pieces exist. The research is done. The code works. What was missing was the intelligence to assemble them. That's what OpenCLAW is for."*
|
|
|