Official a2a switch

examples/nanda_agent.py

@@ -1,273 +1,156 @@
+# examples/nanda_agent.py
 #!/usr/bin/env python3
 """
-LLM-Powered Modular NANDA Agent
+LLM-Powered NANDA Agent
 
-This agent uses Anthropic Claude for intelligent responses based on configurable personality and expertise.
-Simply update the AGENT_CONFIG section to create different agent personalities.
+Configurable agent using Anthropic Claude for intelligent responses.
+Customize via environment variables or the config dict.
 """
 import os
 import sys
-import time
+import asyncio
 import uuid
 from datetime import datetime
-from typing import Dict, List, Any
 
-# Add the parent directory to the path to allow importing streamlined_adapter
 sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
 
 from nanda_core.core.adapter import NANDA
 
-# Try to import Anthropic - will fail gracefully if not available
+from dotenv import load_dotenv
+load_dotenv()
+
 try:
     from anthropic import Anthropic
     ANTHROPIC_AVAILABLE = True
 except ImportError:
     ANTHROPIC_AVAILABLE = False
-    print("⚠️ Warning: anthropic library not available. Install with: pip install anthropic")
+    print("⚠️ anthropic library not available. Install with: pip install anthropic")
 
 # =============================================================================
-# AGENT CONFIGURATION - Customize this section for different agents
+# CONFIGURATION
 # =============================================================================
 
-# Get configuration from environment variables or use defaults
-def get_agent_config():
-    """Load agent configuration from environment variables or use defaults"""
+def get_config():
+    """Load configuration from environment variables"""
+    base_id = os.getenv("AGENT_ID", "helpful-agent")
+    agent_id = f"{base_id}-{uuid.uuid4().hex[:6]}" if '-' not in base_id else base_id
 
-    # Generate agent_id with hex suffix for uniqueness
-    base_agent_id = os.getenv("AGENT_ID", "helpful-ubuntu-agent")
-    if not base_agent_id.endswith('-') and '-' not in base_agent_id.split('-')[-1]:
-        # Add 6-character hex suffix if not already present
-        hex_suffix = uuid.uuid4().hex[:6]
-        agent_id = f"{base_agent_id}-{hex_suffix}"
-    else:
-        agent_id = base_agent_id
+    capabilities = os.getenv("AGENT_CAPABILITIES", "general assistance,conversation")
+    capabilities_list = [cap.strip() for cap in capabilities.split(",")]
 
-    print(f"Generated agent_id: {agent_id}")
-    agent_name = os.getenv("AGENT_NAME", "Ubuntu Helper")
+    agent_name = os.getenv("AGENT_NAME", "Helper Agent")
     domain = os.getenv("AGENT_DOMAIN", "general assistance")
-    specialization = os.getenv("AGENT_SPECIALIZATION", "helpful and friendly AI assistant")
-    description = os.getenv("AGENT_DESCRIPTION", "I am a helpful AI assistant specializing in general tasks and Ubuntu system administration.")
-    capabilities = os.getenv("AGENT_CAPABILITIES", "general assistance,Ubuntu system administration,Python development,cloud deployment,agent-to-agent communication")
-    registry_url = os.getenv("REGISTRY_URL", None)
-    public_url = os.getenv("PUBLIC_URL", None)
-
-    # Parse capabilities into a list
-    expertise_list = [cap.strip() for cap in capabilities.split(",")]
+    specialization = os.getenv("AGENT_SPECIALIZATION", "helpful AI assistant")
+    description = os.getenv("AGENT_DESCRIPTION", "I'm a helpful AI assistant.")
 
-    # Create dynamic system prompt based on configuration
-    system_prompt = f"""You are {agent_name}, a {specialization} working in the domain of {domain}.
+    system_prompt = f"""You are {agent_name}, a {specialization} in {domain}.
 
 {description}
 
-You are part of the NANDA (Network of Autonomous Distributed Agents) system. You can communicate with other agents and help users with various tasks.
+You are part of the NANDA agent network and can communicate with other agents using @agent-id syntax.
 
-Your capabilities include:
-{chr(10).join([f"- {cap}" for cap in expertise_list])}
+Your capabilities: {', '.join(capabilities_list)}
 
-Always be helpful, accurate, and concise in your responses. If you're unsure about something, say so honestly. You can also help with basic calculations, provide time information, and engage in casual conversation.
-
-When someone asks about yourself, mention that you're part of the NANDA agent network and can communicate with other agents using the @agent_name syntax."""
+Be helpful, accurate, and concise."""
 
     return {
         "agent_id": agent_id,
         "agent_name": agent_name,
         "domain": domain,
         "specialization": specialization,
         "description": description,
-        "expertise": expertise_list,
-        "registry_url": registry_url,
-        "public_url": public_url,
-        "system_prompt": system_prompt,
+        "capabilities": capabilities_list,
+        "port": int(os.getenv("PORT", "6000")),
+        "registry_url": os.getenv("REGISTRY_URL"),
+        "public_url": os.getenv("PUBLIC_URL") or f"http://localhost:{os.getenv('PORT', '6000')}",
         "anthropic_api_key": os.getenv("ANTHROPIC_API_KEY"),
-        "model": "claude-3-haiku-20240307"  # Fast and cost-effective model
+        "model": os.getenv("ANTHROPIC_MODEL", "claude-3-haiku-20240307"),
+        "system_prompt": system_prompt
     }
 
-# Load configuration
-AGENT_CONFIG = get_agent_config()
-
-# Port configuration - use environment variable or default to 6000
-PORT = int(os.getenv("PORT", "6000"))
-
 # =============================================================================
-# LLM-POWERED AGENT LOGIC - Uses Anthropic Claude for intelligent responses
+# AGENT LOGIC
 # =============================================================================
 
-def create_llm_agent_logic(config: Dict[str, Any]):
-    """
-    Creates an LLM-powered agent logic function based on the provided configuration.
-    Uses Anthropic Claude for intelligent, context-aware responses.
-    """
+def create_agent_logic(config):
+    """Create agent logic function with LLM or fallback"""
 
     # Initialize Anthropic client
-    anthropic_client = None
+    client = None
     if ANTHROPIC_AVAILABLE and config.get("anthropic_api_key"):
         try:
-            anthropic_client = Anthropic(api_key=config["anthropic_api_key"])
-            print(f"✅ Anthropic Claude initialized for {config['agent_name']}")
+            client = Anthropic(api_key=config["anthropic_api_key"])
+            print(f"✅ Claude initialized ({config['model']})")
         except Exception as e:
-            print(f"❌ Failed to initialize Anthropic: {e}")
-            anthropic_client = None
-
-    # Prepare system prompt (already formatted in get_agent_config)
-    system_prompt = config["system_prompt"]
+            print(f"❌ Claude initialization failed: {e}")
 
-    def llm_agent_logic(message: str, conversation_id: str) -> str:
-        """LLM-powered agent logic with fallback to basic responses"""
+    def agent_logic(message: str, conversation_id: str) -> str:
+        """Process message with Claude or fallback"""
 
-        # If LLM is available, use it for intelligent responses
-        if anthropic_client:
+        if client:
             try:
-                # Add current time context if time-related query
-                context_info = ""
-                if any(time_word in message.lower() for time_word in ['time', 'date', 'when']):
-                    context_info = f"\n\nCurrent time: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}"
+                # Add time context if relevant
+                context = ""
+                if any(word in message.lower() for word in ['time', 'date', 'when']):
+                    context = f"\n\nCurrent time: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}"
 
-                response = anthropic_client.messages.create(
+                response = client.messages.create(
                     model=config["model"],
                     max_tokens=500,
-                    system=system_prompt + context_info,
-                    messages=[
-                        {
-                            "role": "user", 
-                            "content": message
-                        }
-                    ]
+                    system=config["system_prompt"] + context,
+                    messages=[{"role": "user", "content": message}]
                 )
 
                 return response.content[0].text.strip()
 
             except Exception as e:
-                print(f"❌ LLM Error: {e}")
-                # Fall back to basic response
-                return f"Sorry, I'm having trouble processing that right now. Error: {str(e)}"
+                return f"Error: {str(e)}"
 
-        # Fallback to basic responses if LLM not available
+        # Fallback responses
+        msg = message.lower().strip()
+        if any(greeting in msg for greeting in ['hello', 'hi', 'hey']):
+            return f"Hello! I'm {config['agent_name']}. Set ANTHROPIC_API_KEY for full capabilities."
+        elif 'time' in msg:
+            return f"Current time: {datetime.now().strftime('%H:%M:%S')}"
         else:
-            return _basic_fallback_response(message, config)
-
-    return llm_agent_logic
-
-def _basic_fallback_response(message: str, config: Dict[str, Any]) -> str:
-    """Basic fallback responses when LLM is not available"""
-    msg = message.lower().strip()
-
-    # Handle greetings
-    if any(greeting in msg for greeting in ['hello', 'hi', 'hey']):
-        return f"Hello! I'm {config['agent_name']}, but I need an Anthropic API key to provide intelligent responses. Please set ANTHROPIC_API_KEY environment variable."
+            return f"I'm {config['agent_name']}. Set ANTHROPIC_API_KEY to enable LLM responses."
 
-    # Handle time requests
-    elif 'time' in msg:
-        current_time = datetime.now().strftime("%H:%M:%S")
-        return f"The current time is {current_time}."
-
-    # Handle basic calculations
-    elif any(op in message for op in ['+', '-', '*', '/', '=']):
-        try:
-            calculation = message.replace('x', '*').replace('X', '*').replace('=', '').strip()
-            result = eval(calculation)
-            return f"Calculation result: {calculation} = {result}"
-        except:
-            return "Sorry, I couldn't calculate that. Please check your expression."
-
-    # Default fallback
-    else:
-        return f"I'm {config['agent_name']}, but I need an Anthropic API key to provide intelligent responses. Please set ANTHROPIC_API_KEY environment variable and restart me."
+    return agent_logic
 
 # =============================================================================
-# MAIN EXECUTION
+# MAIN
 # =============================================================================
 
-def main():
-    """Main function to start the LLM-powered modular agent"""
-    print(f"🤖 Starting {AGENT_CONFIG['agent_name']}")
-    print(f"📝 Specialization: {AGENT_CONFIG['specialization']}")
-    print(f"🎯 Domain: {AGENT_CONFIG['domain']}")
-    print(f"🛠️ Capabilities: {', '.join(AGENT_CONFIG['expertise'])}")
-    if AGENT_CONFIG['registry_url']:
-        print(f"🌐 Registry: {AGENT_CONFIG['registry_url']}")
+async def main():
+    config = get_config()
 
-    # Check for Anthropic API key
-    if not AGENT_CONFIG.get("anthropic_api_key"):
-        print("⚠️ Warning: ANTHROPIC_API_KEY not found in environment variables")
-        print("   The agent will use basic fallback responses only")
-        print("   Set ANTHROPIC_API_KEY to enable LLM capabilities")
-    else:
-        print(f"🧠 LLM Model: {AGENT_CONFIG['model']}")
+    print(f"🤖 {config['agent_name']} ({config['agent_id']})")
+    print(f"🎯 {config['domain']} - {config['specialization']}")
+    print(f"🔗 {config['public_url']}/a2a")
+    if config['registry_url']:
+        print(f"🌐 Registry: {config['registry_url']}")
 
-    # Create the LLM-powered agent logic based on configuration
-    agent_logic = create_llm_agent_logic(AGENT_CONFIG)
+    agent_logic = create_agent_logic(config)
 
-    # Create and start the NANDA agent
     nanda = NANDA(
-        agent_id=AGENT_CONFIG["agent_id"],
+        agent_id=config["agent_id"],
         agent_logic=agent_logic,
-        port=PORT,
-        registry_url=AGENT_CONFIG["registry_url"],
-        public_url=AGENT_CONFIG["public_url"],
-        enable_telemetry=True
+        agent_name=config["agent_name"],
+        domain=config["domain"],
+        specialization=config["specialization"],
+        description=config["description"],
+        capabilities=config["capabilities"],
+        port=config["port"],
+        registry_url=config["registry_url"],
+        public_url=config["public_url"],
+        enable_telemetry=True,
+        protocols={"a2a": {"enabled": True}}
     )
 
-    print(f"🚀 Agent URL: http://localhost:{PORT}/a2a")
-    print("💡 Try these messages:")
-    print("   - 'Hello there'")
-    print("   - 'Tell me about yourself'")
-    print("   - 'What time is it?'")
-    print("   - 'How can you help with Ubuntu?'")
-    print("   - 'Explain Python virtual environments'")
-    print("   - '5 + 3'")
-    print("\n🛑 Press Ctrl+C to stop")
-
-    # Start the agent
-    nanda.start()
-
-def create_custom_agent(agent_name, specialization, domain, expertise_list, port=6000, anthropic_api_key=None, registry_url=None):
-    """
-    Helper function to quickly create a custom LLM-powered agent with different config
-
-    Example usage:
-        create_custom_agent(
-            agent_name="Data Scientist", 
-            specialization="analytical and precise AI assistant",
-            domain="data science",
-            expertise_list=["data analysis", "statistics", "machine learning", "Python"],
-            port=6001,
-            anthropic_api_key="sk-ant-xxxxx"
-        )
-    """
-    custom_config = AGENT_CONFIG.copy()
-    custom_config.update({
-        "agent_id": agent_name.lower().replace(" ", "-"),
-        "agent_name": agent_name,
-        "specialization": specialization,
-        "domain": domain,
-        "expertise": expertise_list,
-        "registry_url": registry_url,
-        "anthropic_api_key": anthropic_api_key or os.getenv("ANTHROPIC_API_KEY"),
-        "system_prompt": f"""You are {agent_name}, a {specialization} working in the domain of {domain}. 
-
-You are part of the NANDA (Network of Autonomous Distributed Agents) system. You can communicate with other agents and help users with various tasks.
-
-Your capabilities include:
-{chr(10).join([f"- {expertise}" for expertise in expertise_list])}
-
-Always be helpful, accurate, and concise in your responses. If you're unsure about something, say so honestly.
-
-When someone asks about yourself, mention that you're part of the NANDA agent network and can communicate with other agents using the @agent_name syntax."""
-    })
-
-    agent_logic = create_llm_agent_logic(custom_config)
-
-    nanda = NANDA(
-        agent_id=custom_config["agent_id"],
-        agent_logic=agent_logic,
-        port=port,
-        registry_url=custom_config["registry_url"],
-        enable_telemetry=True
-    )
+    print("\n💬 Try: 'Hello', 'What time is it?', '@other-agent Hello!'")
+    print("🛑 Press Ctrl+C to stop\n")
 
-    print(f"🤖 Starting custom LLM agent: {agent_name}")
-    print(f"🚀 Agent URL: http://localhost:{port}/a2a")
-    nanda.start()
+    await nanda.start()
 
 if __name__ == "__main__":
-    main()
+    asyncio.run(main())

nanda_core/__init__.py

@@ -4,9 +4,9 @@
 """
 
 from .core.adapter import NANDA
-from .core.agent_bridge import SimpleAgentBridge
+from .core.agent_bridge import AgentBridge
 
 __all__ = [
     "NANDA",
-    "SimpleAgentBridge"
+    "AgentBridge"
 ]

nanda_core/core/__init__.py

@@ -4,9 +4,9 @@
 """
 
 from .adapter import NANDA
-from .agent_bridge import SimpleAgentBridge
+from .agent_bridge import AgentBridge
 
 __all__ = [
     "NANDA",
-    "SimpleAgentBridge"
+    "AgentBridge"
 ]