AWS MCP Server

Instructions

Implementation Reference

• src/mcp_server_aws/server.py:165-167 (handler)

  Core execution logic for the s3_object_list tool: calls boto3 S3 client's list_objects_v2 method with the bucket_name argument.

  elif name == "s3_object_list":
      response = s3_client.list_objects_v2(
          Bucket=arguments["bucket_name"])

• src/mcp_server_aws/tools.py:82-95 (schema)

  Tool definition including input schema that requires a 'bucket_name' string parameter.

  Tool(
      name="s3_object_list",
      description="List objects in an S3 bucket",
      inputSchema={
          "type": "object",
          "properties": {
              "bucket_name": {
                  "type": "string",
                  "description": "Name of the S3 bucket"
              }
          },
          "required": ["bucket_name"]
      }
  ),

• src/mcp_server_aws/server.py:135-139 (registration)

  Registers the s3_object_list tool (among others) by returning the list of AWS tools via MCP's list_tools handler.

  @server.list_tools()
  async def list_tools() -> list[Tool]:
      """List available AWS tools"""
      logger.debug("Handling list_tools request")
      return get_aws_tools()

• src/mcp_server_aws/server.py:178-180 (handler)

  Shared post-execution logic: logs the operation to audit and returns the S3 API response as formatted JSON text content.

  aws.log_operation("s3", name.replace("s3_", ""), arguments)
  return [TextContent(type="text", text=f"Operation Result:\n{json.dumps(response, indent=2, default=custom_json_serializer)}")]

• src/mcp_server_aws/tools.py:450-454 (registration)

  Combines S3 and DynamoDB tools (including s3_object_list from get_s3_tools) into the full tools list returned by list_tools.

  def get_aws_tools() -> list[Tool]:
      return [
          *get_s3_tools(),
          *get_dynamodb_tools()
      ]