Jira MCP Server

import os import logging # Configure logging to stderr so it doesn't interfere with MCP protocol logging.basicConfig( level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s', handlers=[logging.StreamHandler()] ) # MCP Configuration MCP_TRANSPORT = os.environ.get("MCP_TRANSPORT", "stdio") # Snowflake API configuration from environment variables SNOWFLAKE_BASE_URL = os.environ.get("SNOWFLAKE_BASE_URL") SNOWFLAKE_DATABASE = os.environ.get("SNOWFLAKE_DATABASE") SNOWFLAKE_SCHEMA = os.environ.get("SNOWFLAKE_SCHEMA") # Snowflake token handling - for stdio transport, get from environment # For other transports, it will be retrieved from request context in tools layer if MCP_TRANSPORT == "stdio": SNOWFLAKE_TOKEN = os.environ.get("SNOWFLAKE_TOKEN") else: # For non-stdio transports, token will be passed from tools layer SNOWFLAKE_TOKEN = None # Prometheus metrics configuration ENABLE_METRICS = os.environ.get("ENABLE_METRICS", "false").lower() == "true" METRICS_PORT = int(os.environ.get("METRICS_PORT", "8000")) # Check if Prometheus is available try: from prometheus_client import Counter, Histogram, Gauge, CONTENT_TYPE_LATEST, generate_latest PROMETHEUS_AVAILABLE = True except ImportError: PROMETHEUS_AVAILABLE = False