Personal Resume Agent

# Personal Resume Agent A personalized AI agent that reads your resume and provides intelligent responses about your professional background. Supports both local MCP (Model Context Protocol) integration with Claude Desktop and deployment to the NANDA (Network of Autonomous Decentralized Agents) for Agent-to-Agent (A2A) communication. ## Features ### Core Capabilities - **Resume Processing**: Automatically reads and processes resume files (PDF, DOCX, TXT, MD) - **RAG System**: Uses ChromaDB and sentence transformers for intelligent content retrieval - **MCP Server**: Exposes functionality through standardized MCP protocol for Claude Desktop - **Skill Matching**: Analyzes how well your skills match job requirements - **Natural Language Interface**: Ask questions about your experience, skills, education, etc. ### NANDA Network Integration - **A2A Communication**: Connect with 484+ agents in the NANDA network - **Agent-to-Agent Messaging**: Your resume agent can communicate with other specialized agents - **AWS Deployment**: Production-ready deployment on EC2 with supervisor - **Interactive Dashboards**: CLI tools for sending tasks, viewing messages, and chatting with agents - **Network Discovery**: Registered in NANDA registry for agent discoverability ## Quick Start ### Local Development (MCP Server) 1. **Install Dependencies** ```bash pip install -r requirements.txt ``` 2. **Add Your Resume** ```bash # Place your resume files in the data/ directory cp your-resume.pdf data/ ``` 3. **Test the Agent** ```bash cd src python personal_resume_agent.py ``` 4. **Run as MCP Server** ```bash cd src python mcp_resume_server.py ``` ### NANDA Network Deployment (A2A) 1. **Deploy to AWS EC2** ```bash # Configure deployment script with your details ./deploy_aws_nest.sh ``` 2. **Upload Resume Files** ```bash ./upload_resume.sh ``` 3. **Verify Deployment** ```bash # Check agent status ssh ubuntu@<your-ec2-ip> "sudo supervisorctl status resume-agent" ``` 4. **Use Interactive Dashboards** ```bash # Send tasks to your agent python3 send_task.py # View incoming messages python3 view_messages.py --monitor # Chat with other agents in NANDA network python3 agent_chat.py ``` ## Project Structure ``` personal-resume-agent/ ├── src/ # Core source code │ ├── resume_rag.py # RAG system (ChromaDB + embeddings) │ ├── personal_resume_agent.py # Main agent logic │ └── mcp_resume_server.py # MCP server for Claude Desktop │ ├── NANDA Integration │ ├── nest_resume_agent.py # NANDA-compatible agent wrapper │ ├── deploy_aws_nest.sh # AWS EC2 deployment script │ └── upload_resume.sh # Resume file upload script │ ├── Interactive Dashboards │ ├── send_task.py # Send tasks to resume agent │ ├── view_messages.py # View incoming A2A messages │ └── agent_chat.py # Chat with other NANDA agents │ ├── Documentation │ ├── README.md # This file │ ├── DASHBOARD_GUIDE.md # Dashboard usage guide │ ├── A2A_CHAT_GUIDE.md # Agent-to-Agent chat guide │ ├── A2A_GUIDE.md # A2A protocol documentation │ └── MESSAGE_DASHBOARD.md # Message viewing guide │ ├── data/ # Resume files (gitignored) ├── tests/ # Test files └── requirements.txt # Python dependencies ``` ## Usage Examples ### Direct Agent Usage ```python from personal_resume_agent import PersonalResumeAgent agent = PersonalResumeAgent() await agent.initialize() # Ask questions about your resume result = await agent.process_query("What programming languages do I know?") print(result['response']) # Analyze skill match for a job match = await agent.get_skill_match("Python, React, AWS, Docker") print(f"Match: {match['match_percentage']}%") ``` ### MCP Server Tools The MCP server exposes these tools: - `query_resume`: Ask questions about resume content - `get_agent_info`: Get agent capabilities and status - `analyze_skill_match`: Compare skills with job requirements - `get_resume_summary`: Get overview of resume knowledge base ### Interactive Dashboards #### 1. Send Task Dashboard (`send_task.py`) Interactive CLI for sending queries to your resume agent: ```bash # Interactive mode python3 send_task.py # Direct query python3 send_task.py "What are my technical skills?" ``` **Features:** - Color-coded interface - Session statistics tracking - Conversation history (last 10) - Loading animations - Special commands: `/help`, `/history`, `/stats`, `/clear` #### 2. Message Viewer (`view_messages.py`) Monitor incoming A2A messages from other agents: ```bash # View recent messages python3 view_messages.py # Real-time monitoring mode python3 view_messages.py --monitor # Show statistics python3 view_messages.py --stats # Filter by sender python3 view_messages.py --filter tech-expert ``` **Features:** - Real-time message monitoring - Sender statistics with bar charts - Message filtering - Auto-refresh capability #### 3. A2A Network Chat (`agent_chat.py`) Connect with 484+ agents in NANDA network: ```bash python3 agent_chat.py ``` **Commands:** - `/list` - Show all available agents - `/search <keyword>` - Find agents by expertise - `/info <agent-id>` - View agent details - `/chat <agent-id>` - Start chatting with an agent - `/history` - View conversation history - `/help` - Show all commands **Example session:** ``` 📝 > /list 📝 > /chat tech-expert 💬 [tech-expert] > What are the key technical skills? 💬 [tech-expert] > back 📝 > /history ``` ## NANDA Network Deployment ### Prerequisites - AWS account with EC2 access - SSH key pair for EC2 instances - NANDA framework installed locally ### Step-by-Step Deployment #### 1. Configure Deployment Script Edit `deploy_aws_nest.sh` with your details: ```bash KEY_NAME="your-key-name" KEY_FILE="/path/to/your-key.pem" SECURITY_GROUP_ID="sg-xxxxxxxxx" ``` #### 2. Deploy to AWS EC2 ```bash chmod +x deploy_aws_nest.sh ./deploy_aws_nest.sh ``` This script will: - Launch t3.small EC2 instance (20GB disk, CPU-optimized) - Install Python 3.11 and dependencies - Install PyTorch CPU-only version (saves ~2GB space) - Clone repository from GitHub - Configure supervisor for process management - Register agent in NANDA registry #### 3. Upload Resume Files ```bash chmod +x upload_resume.sh ./upload_resume.sh ``` #### 4. Verify Deployment ```bash # Check agent status ssh -i your-key.pem ubuntu@<ec2-ip> "sudo supervisorctl status resume-agent" # View logs ssh -i your-key.pem ubuntu@<ec2-ip> "sudo tail -f /var/log/resume-agent.log" # Test agent endpoint curl http://<ec2-ip>:6050/a2a \ -H 'Content-Type: application/json' \ -d '{"content": {"text": "What are my skills?", "type": "text"}, "role": "user", "conversation_id": "test-001"}' ``` ### NANDA Network Information Once deployed, your agent will be: - **Registered** in NANDA registry (http://registry.chat39.com:6900) - **Discoverable** by 484+ other agents in the network - **Accessible** via A2A protocol endpoint - **Visible** on NANDA dashboard (https://index.projectnanda.org) ### Architecture: NANDA Deployment ``` ┌─────────────────────────────────────────────────────────────┐ │ NANDA Network │ │ (484+ Connected Agents) │ └──────────────────────┬──────────────────────────────────────┘ │ │ A2A Protocol │ ┌──────────────────────▼──────────────────────┐ │ AWS EC2 (t3.small) │ │ ┌────────────────────────────┐ │ │ │ Supervisor Process Mgr │ │ │ └──────────┬─────────────────┘ │ │ │ │ │ ┌──────────▼─────────────────┐ │ │ │ NANDA Adapter │ │ │ │ (nest_resume_agent.py) │ │ │ └──────────┬─────────────────┘ │ │ │ │ │ ┌──────────▼─────────────────┐ │ │ │ Resume Agent + RAG │ │ │ │ (ChromaDB + Embeddings) │ │ │ └────────────────────────────┘ │ │ │ │ Port: 6050/a2a │ │ Registry: registry.chat39.com:6900 │ └─────────────────────────────────────────────┘ ``` ### Environment Variables Set these in your deployment: ```bash AGENT_ID=resume-agent # Your agent identifier PORT=6050 # HTTP server port REGISTRY_URL=http://registry.chat39.com:6900 # NANDA registry PUBLIC_URL=http://<ec2-ip>:6050 # Your agent's public URL DATA_DIR=/home/resume/resume-agent/data # Resume files directory ``` ## Configuration ### Claude Desktop Integration Add to your Claude Desktop config (`claude_desktop_config.json`): ```json { "mcpServers": { "personal-resume": { "command": "python", "args": ["/path/to/personal-resume-agent/src/mcp_resume_server.py"], "cwd": "/path/to/personal-resume-agent" } } } ``` ## Supported File Formats - **PDF**: Extracted using PyPDF2 - **DOCX**: Processed with python-docx - **TXT/MD**: Plain text files ## Requirements - Python 3.8+ - ChromaDB for vector storage - Sentence Transformers for embeddings - PyPDF2 for PDF processing - python-docx for Word documents ## Privacy & Security 🔒 **Important Privacy Notes:** - All resume data is processed **locally** on your machine - No personal information is sent to external services - Vector database is stored locally in `data/resume_vectordb/` - The `data/` directory is excluded from version control - Never commit personal resume files to public repositories ## Architecture ### Local MCP Mode ``` ┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐ │ Resume Files │───▶│ RAG System │───▶│ MCP Server │ │ (PDF/DOCX) │ │ (ChromaDB + │ │ (Claude Tool) │ │ │ │ Transformers) │ │ │ └─────────────────┘ └─────────────────┘ └─────────────────┘ │ ▼ ┌─────────────────┐ │ Personal Resume │ │ Agent │ │ (Query Engine) │ └─────────────────┘ │ ▼ ┌─────────────────┐ │ Claude Desktop │ └─────────────────┘ ``` ### NANDA Network Mode ``` ┌────────────────────────────────────────────────────────────┐ │ NANDA Network (484+ Agents) │ │ ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐ │ │ │tech- │ │python- │ │financial-│ │ai- │ │ │ │expert │ │expert │ │advisor │ │researcher│ │ │ └────┬─────┘ └────┬─────┘ └────┬─────┘ └────┬─────┘ │ │ │ │ │ │ │ └───────┼─────────────┼──────────────┼─────────────┼─────────┘ │ │ │ │ └─────────────┴──────────────┴─────────────┘ │ A2A Protocol ▼ ┌────────────────────────┐ │ NANDA Adapter │ │ (nest_resume_agent.py) │ └────────────┬───────────┘ │ ┌────────────▼───────────┐ │ Resume Agent + RAG │ │ (ChromaDB + LLM) │ └────────────┬───────────┘ │ ┌────────────▼───────────┐ │ Resume Data Store │ │ (Vector Database) │ └────────────────────────┘ Accessible via: - A2A Endpoint: http://<ec2-ip>:6050/a2a - Registry: http://registry.chat39.com:6900 - Dashboard: https://index.projectnanda.org ``` ## Documentation Comprehensive guides are available: - **[DASHBOARD_GUIDE.md](DASHBOARD_GUIDE.md)** - Interactive dashboard usage - Send Task Dashboard - Message Viewer - Command reference and examples - **[A2A_CHAT_GUIDE.md](A2A_CHAT_GUIDE.md)** - Agent-to-Agent communication - Connecting with 484+ agents - Chat commands and workflows - Use cases and examples - **[A2A_GUIDE.md](A2A_GUIDE.md)** - A2A protocol details - Message format - Protocol specification - Integration patterns - **[MESSAGE_DASHBOARD.md](MESSAGE_DASHBOARD.md)** - Message monitoring - Viewing incoming messages - Log analysis - Real-time monitoring ## Contributing 1. Fork the repository 2. Create a feature branch 3. Make your changes 4. Add tests if applicable 5. Submit a pull request ## License MIT License - See LICENSE file for details.