MCP Business AI Transformation

from fastapi import Request import time import structlog from prometheus_client import Counter, Histogram, Gauge, generate_latest, CONTENT_TYPE_LATEST from fastapi import Response logger = structlog.get_logger() # Prometheus metrics REQUEST_COUNT = Counter( 'http_requests_total', 'Total HTTP requests', ['method', 'endpoint', 'status_code'] ) REQUEST_DURATION = Histogram( 'http_request_duration_seconds', 'HTTP request duration in seconds', ['method', 'endpoint'] ) ACTIVE_CONNECTIONS = Gauge( 'active_connections', 'Number of active connections' ) TOOL_EXECUTIONS = Counter( 'tool_executions_total', 'Total tool executions', ['tool_name', 'status'] ) TOOL_EXECUTION_DURATION = Histogram( 'tool_execution_duration_seconds', 'Tool execution duration in seconds', ['tool_name'] ) AGENT_TASKS = Counter( 'agent_tasks_total', 'Total agent tasks', ['agent_id', 'status'] ) LLM_TOKENS_USED = Counter( 'llm_tokens_used_total', 'Total LLM tokens used', ['provider', 'model'] ) class MetricsMiddleware: """Middleware to collect Prometheus metrics""" async def __call__(self, request: Request, call_next): start_time = time.time() # Increment active connections ACTIVE_CONNECTIONS.inc() try: # Process request response = await call_next(request) # Calculate duration duration = time.time() - start_time # Record metrics REQUEST_COUNT.labels( method=request.method, endpoint=request.url.path, status_code=response.status_code ).inc() REQUEST_DURATION.labels( method=request.method, endpoint=request.url.path ).observe(duration) # Log if slow request if duration > 1.0: logger.warning( "Slow request detected", method=request.method, path=request.url.path, duration=duration, correlation_id=getattr(request.state, "correlation_id", None) ) return response except Exception as e: # Record error metrics REQUEST_COUNT.labels( method=request.method, endpoint=request.url.path, status_code=500 ).inc() logger.error( "Request failed with exception", method=request.method, path=request.url.path, error=str(e), correlation_id=getattr(request.state, "correlation_id", None) ) raise finally: # Decrement active connections ACTIVE_CONNECTIONS.dec() def get_metrics(): """Get Prometheus metrics as HTTP response""" return Response( content=generate_latest(), media_type=CONTENT_TYPE_LATEST ) def record_tool_execution(tool_name: str, success: bool, duration: float): """Record tool execution metrics""" TOOL_EXECUTIONS.labels( tool_name=tool_name, status='success' if success else 'error' ).inc() TOOL_EXECUTION_DURATION.labels(tool_name=tool_name).observe(duration) def record_agent_task(agent_id: str, status: str): """Record agent task metrics""" AGENT_TASKS.labels(agent_id=agent_id, status=status).inc() def record_llm_tokens(provider: str, model: str, tokens: int): """Record LLM token usage""" LLM_TOKENS_USED.labels(provider=provider, model=model).inc(tokens)