Add sse transport

src/fleet_mcp/cli.py

@@ -84,19 +84,33 @@ def _load_config(ctx: click.Context) -> FleetConfig:
         sys.exit(1)
 
 
+
 @cli.command()
+@click.option(
+    "--port",
+    "-p",
+    type=int,
+    default=None,
+    help="Port to listen on for HTTP transports (default: 8000, env: FLEET_MCP_PORT)",
+    show_default=False,
+)
 @click.pass_context
-def run(ctx: click.Context) -> None:
+def run(ctx: click.Context, port: int | None = None) -> None:
     """Run the Fleet MCP server."""
     config = _load_config(ctx)
 
+    # Dynamically determine port: CLI > env > default
+    if port is None:
+        import os
+        port = int(os.environ.get("FLEET_MCP_PORT", 8000))
+
     try:
         server = FleetMCPServer(config)
         readonly_status = " (READ-ONLY MODE)" if config.readonly else ""
         click.echo(
             f"Starting Fleet MCP Server for {config.server_url}{readonly_status}"
         )
-        server.run()
+        server.run(port=port)
     except KeyboardInterrupt:
         click.echo("\nShutting down Fleet MCP Server...")
     except Exception as e:

src/fleet_mcp/config.py

@@ -61,6 +61,11 @@ class FleetConfig(BaseSettings):
         description="Number of hours to retain completed async query results before cleanup",
     )
 
+    transport: str = Field(
+        default="stdio",
+        description="MCP transport protocol: 'stdio', 'sse', or 'streamable-http'",
+    )
+
     @field_validator("server_url")
     @classmethod
     def validate_server_url(cls, v: str) -> str:

src/fleet_mcp/server.py

@@ -1,6 +1,9 @@
 """Fleet MCP Server - Main MCP server implementation."""
 
+
 import logging
+import os
+import argparse
 from typing import Any
 
 from mcp.server.fastmcp import FastMCP
@@ -442,16 +445,24 @@ async def _preload_schema_cache(self) -> None:
             logger.warning(f"Failed to preload schema cache: {e}")
             logger.info("Schema cache will be loaded on first use")
 
-    def run(self) -> None:
-        """Run the MCP server."""
+    def run(self, port: int = 8000) -> None:
+        """Run the MCP server, supporting custom port."""
         import asyncio
+        import uvicorn
 
         logger.info(f"Starting Fleet MCP Server for {self.config.server_url}")
+        logger.info(f"Transport: {self.config.transport}")
+        logger.info(f"Listening on port: {port}")
 
         # Preload schema cache before starting server
         asyncio.run(self._preload_schema_cache())
 
-        self.mcp.run()
+        # Only pass port for HTTP-based transports
+        if self.config.transport in ("sse", "streamable-http"):
+            app = self.mcp.sse_app() if self.config.transport == "sse" else self.mcp.streamable_http_app()
+            uvicorn.run(app, host="0.0.0.0", port=port, log_level="info")
+        else:
+            self.mcp.run(transport=self.config.transport)
 
 
 def create_server(config: FleetConfig | None = None) -> FleetMCPServer:
@@ -467,16 +478,20 @@ def create_server(config: FleetConfig | None = None) -> FleetMCPServer:
 
 
 def main() -> None:
-    """Main entry point for Fleet MCP Server."""
+    """Main entry point for Fleet MCP Server with port support."""
     # Configure logging
     logging.basicConfig(
         level=logging.INFO,
         format="%(asctime)s - %(name)s - %(levelname)s - %(message)s",
     )
 
+    parser = argparse.ArgumentParser(description="Fleet MCP Server")
+    parser.add_argument("--port", "-p", type=int, default=int(os.environ.get("FLEET_MCP_PORT", 8000)), help="Port to listen on for HTTP transports (default: 8000, env: FLEET_MCP_PORT)")
+    args, unknown = parser.parse_known_args()
+
     try:
         server = create_server()
-        server.run()
+        server.run(port=args.port)
     except Exception as e:
         logger.error(f"Failed to start Fleet MCP Server: {e}")
         raise