saap-plattform / backend /agents /colossus_agent.py
Hwandji's picture
feat: initial HuggingFace Space deployment
4343907
#!/usr/bin/env python3
"""
SAAP colossus Server Integration - ColosusSAAPAgent
=================================================
Direct integration with colossus Server für Phase 1 Infrastructure Foundation.
Hybrid Architecture: CachyOS (Orchestrierung) + colossus (LLM Processing)
Server Details:
- URL: https://ai.adrian-schupp.de
- Model: mistral-small3.2:24b-instruct-2506
- Performance Target: < 2s Response-Zeit
Integration with existing SAAP Agent Communication System.
"""
import asyncio
import json
import time
import logging
import os
from typing import Dict, Any, Optional, List
from dataclasses import dataclass, field
import aiohttp
import redis.asyncio as redis
from dotenv import load_dotenv
# Load environment variables
load_dotenv()
# Configure logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
@dataclass
class ColossusConfig:
"""colossus Server Configuration"""
base_url: str = "https://ai.adrian-schupp.de"
api_key: str = field(default_factory=lambda: os.getenv("COLOSSUS_API_KEY", ""))
model: str = "mistral-small3.2:24b-instruct-2506"
max_tokens: int = 1000
temperature: float = 0.7
timeout: int = 30 # seconds
def __post_init__(self):
"""Validate configuration after initialization"""
if not self.api_key:
raise ValueError(
"❌ COLOSSUS_API_KEY environment variable not set.\n"
"Please set it in your .env file:\n"
"COLOSSUS_API_KEY=your-api-key-here"
)
class ColosusSAAPAgent:
"""
SAAP Agent mit colossus Server Integration
Hybrid Architecture:
- CachyOS: Agent Orchestrierung, Message Queue, System Management
- colossus: High-Performance LLM Processing, AI Inference
"""
def __init__(self,
agent_name: str,
agent_role: str = "Coordinator",
config: Optional[ColossusConfig] = None,
redis_url: str = "redis://localhost:6379"):
self.agent_name = agent_name
self.agent_role = agent_role
self.config = config or ColossusConfig()
self.redis_url = redis_url
# Agent context for specialized roles
self.agent_contexts = {
"Coordinator": "Du bist Agent A (Coordinator) für SAAP. Du koordinierst Multi-Agent Workflows und delegierst Tasks effizient.",
"Developer": "Du bist Agent B (Developer) mit Expertise in Python, Node.js, Vue.js. Du fokussierst auf Clean Code und Performance.",
"Analyst": "Du bist Agent C (Analyst) für Requirements-Analyse, Use Cases und Systemdesign. Du lieferst strukturierte Analysen.",
"General": "Du bist ein SAAP Multi-Agent mit genereller KI-Expertise für vielseitige Tasks."
}
# Performance tracking
self.performance_stats = {
"total_requests": 0,
"total_response_time": 0.0,
"average_response_time": 0.0,
"errors": 0,
"successful_requests": 0
}
# Redis connection (will be initialized async)
self.redis_client = None
async def initialize(self):
"""Initialize async components"""
try:
self.redis_client = await redis.from_url(self.redis_url)
await self.redis_client.ping()
logger.info(f"✅ {self.agent_name} connected to Redis")
# Register agent in Redis
await self.register_agent()
except Exception as e:
logger.error(f"❌ Redis connection failed for {self.agent_name}: {e}")
# Continue without Redis - degraded mode
async def register_agent(self):
"""Register agent with SAAP system"""
if not self.redis_client:
return
agent_info = {
"name": self.agent_name,
"role": self.agent_role,
"status": "active",
"model": self.config.model,
"server": "colossus",
"capabilities": ["llm_processing", "multi_agent_communication", "task_coordination"],
"performance_target": "< 2s response time",
"timestamp": time.time()
}
await self.redis_client.hset(
f"agent:{self.agent_name}",
mapping={k: json.dumps(v) if isinstance(v, (dict, list)) else str(v)
for k, v in agent_info.items()}
)
# Add to active agents set
await self.redis_client.sadd("active_agents", self.agent_name)
logger.info(f"📝 {self.agent_name} registered with SAAP system")
async def call_colossus_api(self, prompt: str) -> Dict[str, Any]:
"""
Direct API call to colossus Server
Returns response with performance metrics
"""
start_time = time.time()
try:
headers = {
"Authorization": f"Bearer {self.config.api_key}",
"Content-Type": "application/json"
}
# Add agent context to prompt
context = self.agent_contexts.get(self.agent_role, self.agent_contexts["General"])
enhanced_prompt = f"{context}\\n\\nAufgabe: {prompt}"
payload = {
"model": self.config.model,
"messages": [
{"role": "user", "content": enhanced_prompt}
],
"max_tokens": self.config.max_tokens,
"temperature": self.config.temperature
}
async with aiohttp.ClientSession(timeout=aiohttp.ClientTimeout(total=self.config.timeout)) as session:
async with session.post(
f"{self.config.base_url}/v1/chat/completions",
headers=headers,
json=payload
) as response:
if response.status == 200:
result = await response.json()
# Extract response text
content = result.get("choices", [{}])[0].get("message", {}).get("content", "No response")
# Calculate performance
response_time = time.time() - start_time
# Update stats
self.performance_stats["total_requests"] += 1
self.performance_stats["successful_requests"] += 1
self.performance_stats["total_response_time"] += response_time
self.performance_stats["average_response_time"] = (
self.performance_stats["total_response_time"] /
self.performance_stats["total_requests"]
)
return {
"success": True,
"content": content,
"response_time": response_time,
"model": self.config.model,
"server": "colossus",
"agent": self.agent_name,
"role": self.agent_role,
"performance_check": response_time < 2.0 # Success if < 2s
}
else:
error_text = await response.text()
raise Exception(f"API Error {response.status}: {error_text}")
except Exception as e:
# Update error stats
self.performance_stats["errors"] += 1
self.performance_stats["total_requests"] += 1
response_time = time.time() - start_time
logger.error(f"❌ colossus API call failed: {e}")
return {
"success": False,
"error": str(e),
"response_time": response_time,
"agent": self.agent_name,
"server": "colossus",
"performance_check": False
}
async def process_message(self, message: str, context: Optional[Dict] = None) -> Dict[str, Any]:
"""
Process incoming message with colossus LLM
Integrates with SAAP Message Queue System
"""
# Log message processing
logger.info(f"🤖 {self.agent_name} ({self.agent_role}) processing message...")
# Call colossus API
result = await self.call_colossus_api(message)
# Add SAAP-specific metadata
result.update({
"agent_name": self.agent_name,
"agent_role": self.agent_role,
"timestamp": time.time(),
"message_id": context.get("message_id") if context else None,
"thread_id": context.get("thread_id") if context else None
})
# Store in Redis if available
if self.redis_client and result["success"]:
await self._store_message_result(message, result)
# Performance logging
performance_emoji = "⚡" if result.get("performance_check", False) else "⏱️"
logger.info(f"{performance_emoji} {self.agent_name}: {result.get('response_time', 0):.2f}s")
return result
async def _store_message_result(self, message: str, result: Dict[str, Any]):
"""Store message and result in Redis for monitoring"""
if not self.redis_client:
return
message_data = {
"input": message,
"output": result.get("content", ""),
"agent": self.agent_name,
"role": self.agent_role,
"response_time": result.get("response_time", 0),
"timestamp": result.get("timestamp", time.time()),
"success": result.get("success", False)
}
# Store in message history
await self.redis_client.lpush(
f"messages:{self.agent_name}",
json.dumps(message_data)
)
# Keep only recent messages (last 100)
await self.redis_client.ltrim(f"messages:{self.agent_name}", 0, 99)
# Update agent status
await self.redis_client.hset(
f"agent:{self.agent_name}",
"last_activity",
str(time.time())
)
async def get_performance_stats(self) -> Dict[str, Any]:
"""Get comprehensive performance statistics"""
stats = self.performance_stats.copy()
# Add colossus-specific metrics
stats.update({
"server": "colossus",
"model": self.config.model,
"performance_target_met": stats["average_response_time"] < 2.0,
"success_rate": (
(stats["successful_requests"] / stats["total_requests"]) * 100
if stats["total_requests"] > 0 else 0
),
"agent_name": self.agent_name,
"agent_role": self.agent_role
})
return stats
async def cleanup(self):
"""Cleanup connections"""
if self.redis_client:
await self.redis_client.srem("active_agents", self.agent_name)
await self.redis_client.close()
# Example Usage & Testing
async def test_colossus_integration():
"""Test colossus Server Integration"""
print("🚀 Testing SAAP colossus Server Integration...")
# Create test agents
agents = [
ColosusSAAPAgent("agent_coordinator", "Coordinator"),
ColosusSAAPAgent("agent_developer", "Developer"),
ColosusSAAPAgent("agent_analyst", "Analyst")
]
# Initialize all agents
for agent in agents:
await agent.initialize()
# Test messages
test_messages = [
"Analysiere die SAAP Multi-Agent-Architektur und identifiziere Optimierungsbedarfe.",
"Entwickle Python Code für Redis Message Queue Integration.",
"Erstelle Use Cases für Agent-zu-Agent Kommunikation."
]
# Process messages in parallel
tasks = []
for i, agent in enumerate(agents):
message = test_messages[i % len(test_messages)]
tasks.append(agent.process_message(message, {"test_id": i}))
results = await asyncio.gather(*tasks, return_exceptions=True)
# Print results
print("\\n" + "="*60)
print("🎯 SAAP colossus Integration Results:")
print("="*60)
for i, result in enumerate(results):
if isinstance(result, Exception):
print(f"❌ Agent {i+1}: Error - {result}")
else:
agent_name = result.get("agent_name", f"Agent_{i+1}")
response_time = result.get("response_time", 0)
success = result.get("success", False)
performance = "✅ < 2s" if result.get("performance_check", False) else f"⏱️ {response_time:.2f}s"
print(f"{'✅' if success else '❌'} {agent_name}: {performance}")
if success:
content = result.get("content", "")[:100] + "..." if len(result.get("content", "")) > 100 else result.get("content", "")
print(f" Response: {content}")
# Performance summary
print("\\n" + "="*60)
print("📊 Performance Summary:")
for agent in agents:
stats = await agent.get_performance_stats()
print(f"🤖 {agent.agent_name} ({agent.agent_role}):")
print(f" Average Response Time: {stats['average_response_time']:.2f}s")
print(f" Success Rate: {stats['success_rate']:.1f}%")
print(f" Performance Target Met: {'✅' if stats['performance_target_met'] else '❌'}")
# Cleanup
for agent in agents:
await agent.cleanup()
print("\\n🎉 colossus Integration Test Complete!")
if __name__ == "__main__":
asyncio.run(test_colossus_integration())