app.py

"""

Real MCP Integration - Replace Mock Data with Live Topcoder MCP

This replaces your SimpleIntelligenceEngine with real MCP integration

"""
import asyncio
import httpx
import json
import logging
from typing import List, Dict, Any, Optional
from dataclasses import dataclass, asdict
from datetime import datetime, timedelta

# Configure logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

@dataclass
class Challenge:
    id: str
    title: str
    description: str
    technologies: List[str]
    difficulty: str
    prize: str
    time_estimate: str
    compatibility_score: float = 0.0
    rationale: str = ""

@dataclass
class Skill:
    name: str
    category: str
    description: str
    relevance_score: float = 0.0

@dataclass
class UserProfile:
    skills: List[str]
    experience_level: str
    time_available: str
    interests: List[str]

class RealMCPIntelligenceEngine:
    """Production MCP Integration - Real Topcoder Data"""
    
    def __init__(self):
        self.mcp_url = "https://api.topcoder-dev.com/v6/mcp"
        self.session_id = None
        self.is_connected = False
        self.challenges_cache = {}
        self.skills_cache = {}
        self.cache_expiry = None
        
        # Initialize connection
        asyncio.create_task(self.initialize_connection())
    
    async def initialize_connection(self):
        """Initialize MCP connection and authenticate if needed"""
        try:
            async with httpx.AsyncClient(timeout=30.0) as client:
                
                # Step 1: Try initialization
                init_request = {
                    "jsonrpc": "2.0",
                    "id": 1,
                    "method": "initialize",
                    "params": {
                        "protocolVersion": "2024-11-05",
                        "capabilities": {
                            "roots": {"listChanged": True},
                            "sampling": {}
                        },
                        "clientInfo": {
                            "name": "topcoder-intelligence-assistant",
                            "version": "1.0.0"
                        }
                    }
                }
                
                headers = {
                    "Content-Type": "application/json",
                    "Accept": "application/json, text/event-stream"
                }
                
                response = await client.post(
                    f"{self.mcp_url}/mcp",
                    json=init_request,
                    headers=headers
                )
                
                if response.status_code == 200:
                    result = response.json()
                    if "result" in result:
                        self.is_connected = True
                        logger.info("✅ MCP Connection established")
                        
                        # Extract session info if provided
                        server_info = result["result"].get("serverInfo", {})
                        if "sessionId" in server_info:
                            self.session_id = server_info["sessionId"]
                            logger.info(f"🔑 Session ID obtained: {self.session_id[:10]}...")
                        
                        return True
                
                logger.warning(f"⚠️ MCP initialization failed: {response.status_code}")
                return False
                
        except Exception as e:
            logger.error(f"❌ MCP connection failed: {e}")
            return False
    
    async def call_mcp_tool(self, tool_name: str, arguments: Dict[str, Any]) -> Optional[Dict]:
        """Call an MCP tool with proper error handling"""
        
        if not self.is_connected:
            await self.initialize_connection()
        
        try:
            async with httpx.AsyncClient(timeout=60.0) as client:
                
                request_data = {
                    "jsonrpc": "2.0",
                    "id": datetime.now().timestamp(),
                    "method": "tools/call",
                    "params": {
                        "name": tool_name,
                        "arguments": arguments
                    }
                }
                
                headers = {
                    "Content-Type": "application/json",
                    "Accept": "application/json"
                }
                
                # Add session ID if we have one
                if self.session_id:
                    headers["X-Session-ID"] = self.session_id
                
                response = await client.post(
                    f"{self.mcp_url}/mcp",
                    json=request_data,
                    headers=headers
                )
                
                if response.status_code == 200:
                    result = response.json()
                    if "result" in result:
                        return result["result"]
                    elif "error" in result:
                        logger.error(f"MCP tool error: {result['error']}")
                        return None
                else:
                    logger.error(f"MCP tool call failed: {response.status_code} - {response.text}")
                    return None
                    
        except Exception as e:
            logger.error(f"MCP tool call exception: {e}")
            return None
    
    async def fetch_challenges(self, limit: int = 50, technologies: List[str] = None) -> List[Challenge]:
        """Fetch real challenges from Topcoder MCP"""
        
        # Check cache first
        cache_key = f"challenges_{limit}_{technologies}"
        if (self.cache_expiry and datetime.now() < self.cache_expiry and 
            cache_key in self.challenges_cache):
            return self.challenges_cache[cache_key]
        
        arguments = {"limit": limit}
        if technologies:
            arguments["technologies"] = technologies
        
        result = await self.call_mcp_tool("query-tc-challenges", arguments)
        
        if result and "content" in result:
            challenges_data = result["content"]
            
            challenges = []
            for item in challenges_data:
                if isinstance(item, dict):
                    challenge = Challenge(
                        id=str(item.get("id", "")),
                        title=item.get("title", "Unknown Challenge"),
                        description=item.get("description", "")[:200] + "...",
                        technologies=item.get("technologies", []),
                        difficulty=item.get("difficulty", "Unknown"),
                        prize=f"${item.get('prize', 0):,}",
                        time_estimate=f"{item.get('duration', 0)} hours"
                    )
                    challenges.append(challenge)
            
            # Cache results for 1 hour
            self.challenges_cache[cache_key] = challenges
            self.cache_expiry = datetime.now() + timedelta(hours=1)
            
            logger.info(f"✅ Fetched {len(challenges)} real challenges from MCP")
            return challenges
        
        logger.warning("❌ Failed to fetch challenges, returning empty list")
        return []
    
    async def fetch_skills(self, category: str = None) -> List[Skill]:
        """Fetch real skills from Topcoder MCP"""
        
        cache_key = f"skills_{category}"
        if (self.cache_expiry and datetime.now() < self.cache_expiry and 
            cache_key in self.skills_cache):
            return self.skills_cache[cache_key]
        
        arguments = {}
        if category:
            arguments["category"] = category
        
        result = await self.call_mcp_tool("query-tc-skills", arguments)
        
        if result and "content" in result:
            skills_data = result["content"]
            
            skills = []
            for item in skills_data:
                if isinstance(item, dict):
                    skill = Skill(
                        name=item.get("name", "Unknown Skill"),
                        category=item.get("category", "General"),
                        description=item.get("description", "")
                    )
                    skills.append(skill)
            
            self.skills_cache[cache_key] = skills
            logger.info(f"✅ Fetched {len(skills)} real skills from MCP")
            return skills
        
        logger.warning("❌ Failed to fetch skills, returning empty list")
        return []
    
    def extract_technologies_from_query(self, query: str) -> List[str]:
        """Extract technology keywords from user query"""
        tech_keywords = {
            'python', 'java', 'javascript', 'react', 'node', 'angular', 'vue',
            'aws', 'docker', 'kubernetes', 'api', 'rest', 'graphql', 'sql',
            'mongodb', 'postgresql', 'machine learning', 'ai', 'blockchain',
            'ios', 'android', 'flutter', 'swift', 'kotlin', 'c++', 'c#',
            'ruby', 'php', 'go', 'rust', 'typescript', 'html', 'css'
        }
        
        query_lower = query.lower()
        found_techs = [tech for tech in tech_keywords if tech in query_lower]
        return found_techs
    
    def calculate_compatibility_score(self, challenge: Challenge, user_profile: UserProfile, query: str) -> float:
        """Calculate compatibility score using real challenge data"""
        
        score = 0.0
        factors = []
        
        # 1. Skill matching (40%)
        user_skills_lower = [skill.lower() for skill in user_profile.skills]
        challenge_techs_lower = [tech.lower() for tech in challenge.technologies]
        
        skill_matches = len(set(user_skills_lower) & set(challenge_techs_lower))
        skill_score = min(skill_matches / max(len(challenge.technologies), 1), 1.0) * 0.4
        score += skill_score
        factors.append(f"Skill match: {skill_matches}/{len(challenge.technologies)} technologies")
        
        # 2. Experience level matching (30%)
        experience_mapping = {
            "beginner": {"Beginner": 1.0, "Intermediate": 0.7, "Advanced": 0.3},
            "intermediate": {"Beginner": 0.5, "Intermediate": 1.0, "Advanced": 0.8},
            "advanced": {"Beginner": 0.3, "Intermediate": 0.8, "Advanced": 1.0}
        }
        
        exp_score = experience_mapping.get(user_profile.experience_level.lower(), {}).get(challenge.difficulty, 0.5) * 0.3
        score += exp_score
        factors.append(f"Experience match: {user_profile.experience_level} → {challenge.difficulty}")
        
        # 3. Query relevance (20%)
        query_techs = self.extract_technologies_from_query(query)
        query_matches = len(set([tech.lower() for tech in query_techs]) & set(challenge_techs_lower))
        query_score = min(query_matches / max(len(query_techs), 1), 1.0) * 0.2 if query_techs else 0.1
        score += query_score
        factors.append(f"Query relevance: {query_matches} matches")
        
        # 4. Time availability (10%)
        time_mapping = {
            "2-4 hours": {"1-2 hours": 1.0, "2-4 hours": 1.0, "4+ hours": 0.7},
            "4-8 hours": {"2-4 hours": 0.8, "4+ hours": 1.0, "1-2 hours": 0.6},
            "8+ hours": {"4+ hours": 1.0, "2-4 hours": 0.7, "1-2 hours": 0.4}
        }
        
        time_score = 0.1  # Default
        for user_time, challenge_map in time_mapping.items():
            if user_time in user_profile.time_available:
                time_score = challenge_map.get(challenge.time_estimate, 0.5) * 0.1
                break
        
        score += time_score
        factors.append(f"Time fit: {user_profile.time_available} vs {challenge.time_estimate}")
        
        return min(score, 1.0), factors
    
    async def get_personalized_recommendations(self, user_profile: UserProfile, query: str = "") -> Dict[str, Any]:
        """Get personalized recommendations using real MCP data"""
        
        start_time = datetime.now()
        
        # Fetch real challenges with technology filter if possible
        query_techs = self.extract_technologies_from_query(query)
        challenges = await self.fetch_challenges(limit=100, technologies=query_techs if query_techs else None)
        
        if not challenges:
            # Fallback message
            return {
                "recommendations": [],
                "insights": {
                    "total_challenges": 0,
                    "processing_time": f"{(datetime.now() - start_time).total_seconds():.3f}s",
                    "data_source": "MCP (No data available)",
                    "message": "Unable to fetch real challenge data. Please check MCP connection."
                }
            }
        
        # Score and rank challenges
        scored_challenges = []
        for challenge in challenges:
            score, factors = self.calculate_compatibility_score(challenge, user_profile, query)
            challenge.compatibility_score = score
            challenge.rationale = f"Score: {score:.1%}. " + "; ".join(factors[:2])
            scored_challenges.append(challenge)
        
        # Sort by compatibility score
        scored_challenges.sort(key=lambda x: x.compatibility_score, reverse=True)
        
        # Take top 5 recommendations
        recommendations = scored_challenges[:5]
        
        # Get skills for gap analysis
        skills = await self.fetch_skills()
        
        # Processing time
        processing_time = (datetime.now() - start_time).total_seconds()
        
        return {
            "recommendations": [asdict(rec) for rec in recommendations],
            "insights": {
                "total_challenges": len(challenges),
                "average_score": sum(c.compatibility_score for c in challenges) / len(challenges),
                "processing_time": f"{processing_time:.3f}s",
                "data_source": "Real Topcoder MCP",
                "top_score": recommendations[0].compatibility_score if recommendations else 0,
                "skills_available": len(skills),
                "technologies_detected": query_techs,
                "cache_status": "Fresh data" if not self.cache_expiry else "Cached data"
            }
        }

# Example usage and testing
async def test_real_mcp_engine():
    """Test the real MCP integration"""
    
    print("🚀 Testing Real MCP Integration")
    print("=" * 50)
    
    engine = RealMCPIntelligenceEngine()
    
    # Wait for connection
    await asyncio.sleep(2)
    
    if not engine.is_connected:
        print("❌ MCP connection failed - check authentication")
        return
    
    # Test user profile
    user_profile = UserProfile(
        skills=["Python", "JavaScript", "API"],
        experience_level="Intermediate",
        time_available="4-8 hours",
        interests=["web development", "API integration"]
    )
    
    # Test recommendations
    print("\n🧠 Getting Real Recommendations...")
    recommendations = await engine.get_personalized_recommendations(
        user_profile, 
        "I want to work on Python API challenges"
    )
    
    print(f"\n📊 Results:")
    print(f"   Challenges found: {recommendations['insights']['total_challenges']}")
    print(f"   Processing time: {recommendations['insights']['processing_time']}")
    print(f"   Data source: {recommendations['insights']['data_source']}")
    
    for i, rec in enumerate(recommendations['recommendations'][:3], 1):
        print(f"\n   {i}. {rec['title']}")
        print(f"      Score: {rec['compatibility_score']:.1%}")
        print(f"      Technologies: {', '.join(rec['technologies'][:3])}")

if __name__ == "__main__":
    asyncio.run(test_real_mcp_engine())