AI-powered autonomous agent with memory, decision-making, and modern web interface

An intelligent autonomous agent powered by Claude AI, featuring a dual-memory system, real-time decision-making capabilities, and a beautiful modern web interface built with TypeScript and Tailwind CSS.

• Claude AI Integration - Powered by Anthropic Claude for intelligent decision-making
• Autonomous Agent - Self-directed AI that can set and achieve goals
• Decision Loop Framework - READ → QUERY → THINK → ACT → RECORD → LEARN cycle
• Learning System - Learns from experience and improves over time

• Dual-Layer Architecture
  • Short-term: SQLite database with automatic cleanup
  • Long-term: Qdrant vector database for semantic search
• Auto-cleanup: Maintains optimal memory size automatically
• Type-based organization: Goals, thoughts, actions, observations

• Modern Dashboard - Interactive real-time control panel
• Portfolio Landing Page - Professional showcase with Tailwind CSS
• TypeScript Frontend - Type-safe, maintainable code (700+ lines)
• REST API - 12+ endpoints for complete control
• Real-time Updates - Live status and memory updates

• Figma Integration - Import designs directly from Figma
• Auto-extract - Colors, typography, and components
• Code Generation - Tailwind config and CSS variables
• Component Export - Convert Figma frames to HTML/CSS

• Python 3.11+
• Node.js 18+ (for TypeScript)
• Anthropic API Key (Claude Opus 4.5)

# Clone the repository
git clone https://github.com/yourusername/autonomous-claude.git
cd autonomous-claude

# Install Python dependencies
cd autonomous-claude
pip install -r requirements.txt

# Set up environment variables
export ANTHROPIC_API_KEY='your-anthropic-api-key'
export FIGMA_TOKEN='your-figma-token'  # Optional

# Run the application
python3 web_app.py

• Portfolio: http://localhost:5000/
• Dashboard: http://localhost:5000/dashboard
• Showcase: http://localhost:5000/showcase
• API: http://localhost:5000/api/status

autonomous-claude/
├── autonomous_agent.py      # Main autonomous agent (261 lines)
├── claude_brain.py          # Claude Opus 4.5 integration (332 lines)
├── memory.py                # Memory system (266 lines)
├── decision_loop.py         # Decision framework (202 lines)
├── browser.py               # Browser automation (233 lines)
├── figma_integration.py     # Figma API client (330 lines)
├── web_app.py               # Flask server (350+ lines)
├── templates/
│   ├── index.html           # Dashboard UI
│   ├── portfolio.html       # Landing page
│   └── showcase.html        # Feature showcase
├── static/
│   ├── css/                 # Styles
│   ├── js/                  # Compiled JavaScript
│   └── ts/                  # TypeScript source (700+ lines)
├── tests/
│   ├── test_memory.py       # Memory tests
│   └── test_api.py          # API tests
└── requirements.txt         # Dependencies

POST /api/agent/goal          # Set agent goal
POST /api/agent/start         # Start autonomous agent
POST /api/agent/stop          # Stop agent
GET  /api/status             # Get current status

GET  /api/memories/recent    # Get recent memories
GET  /api/memories/search    # Search by type
GET  /api/memories/stats     # Memory statistics
POST /api/memories/add       # Add new memory
POST /api/database/clear     # Clear all memories

POST /api/chatgpt/think      # Ask ChatGPT to think
POST /api/chatgpt/analyze    # Analyze and decide
POST /api/chatgpt/learn      # Learn from experience

POST /api/figma/import       # Import Figma design

• Python 3.11+ - Core language
• Flask 3.0 - Web framework
• SQLite - Short-term memory database
• Qdrant - Vector database for semantic search
• OpenAI API - ChatGPT integration

• TypeScript 5.0+ - Type-safe JavaScript
• Tailwind CSS - Utility-first styling
• Vanilla JS - No framework dependencies
• Modern CSS - Gradients, animations, responsive design

• OpenAI GPT-3.5/4 - Language model
• Sentence Transformers - Text embeddings
• Vector Search - Semantic similarity

• Figma API - Design import
• REST API - Integration-ready
• CORS - Cross-origin support
• n8n Ready - Workflow automation

# Run all tests
pytest

# Run with coverage
pytest --cov=. --cov-report=html

# Run specific test file
pytest tests/test_memory.py -v

• ✅ Memory system tests
• ✅ API endpoint tests
• ✅ Integration tests
• ✅ Error handling tests

agent.set_goal("Research the latest AI developments")
agent.run()
# Agent autonomously searches, analyzes, and summarizes

from figma_integration import FigmaClient

client = FigmaClient()
design = client.get_file('your-file-key')
colors = client.extract_colors(design)
# Auto-generate design system

brain = ChatGPTBrain()
decision = brain.think("How should I approach this problem?")
# Get intelligent recommendations

# Required
OPENAI_API_KEY=sk-...              # OpenAI API key

# Optional
FIGMA_TOKEN=figd_...               # Figma Personal Access Token
FLASK_ENV=development              # Flask environment
FLASK_APP=web_app.py               # Flask app entry point

Edit memory.py to adjust:

• Memory retention limit (default: 50 entries)
• Database path
• Qdrant connection settings

• Web Interface Guide
• Figma Integration
• ChatGPT Setup
• API Documentation

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

This project is licensed under the MIT License - see the LICENSE file for details.

Built with ❤️ using Claude AI, TypeScript, and Python

• OpenAI for ChatGPT API
• Figma for design integration API
• Qdrant for vector database
• Tailwind CSS for modern styling

• Documentation: View Docs
• Demo: Live Demo
• GitHub: Repository
• Issues: Report Bug

Made with ⚡ by [Your Name]