diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/.python-version b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/.python-version
new file mode 100644
index 0000000..24ee5b1
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/.python-version
@@ -0,0 +1 @@
+3.13
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/README.md b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/README.md
new file mode 100644
index 0000000..e69de29
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/__pycache__/main.cpython-312.pyc b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/__pycache__/main.cpython-312.pyc
new file mode 100644
index 0000000..1bb0314
Binary files /dev/null and b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/__pycache__/main.cpython-312.pyc differ
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/__pycache__/schema_converter.cpython-312.pyc b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/__pycache__/schema_converter.cpython-312.pyc
new file mode 100644
index 0000000..8449988
Binary files /dev/null and b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/__pycache__/schema_converter.cpython-312.pyc differ
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/businessQueries.json b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/businessQueries.json
new file mode 100644
index 0000000..9e90f97
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/businessQueries.json
@@ -0,0 +1,80 @@
+[
+  {
+    "id": "2000851983567065089",
+    "createBy": "bjt",
+    "createTime": "2025-12-16 16:54:22",
+    "updateBy": null,
+    "updateTime": null,
+    "serviceId": "2000821852159709186",
+    "uniqueName": "按摘要查凭证",
+    "name": "anzhaiyaochapingzheng_91da19f8",
+    "description": "按照摘要描述查询凭证信息",
+    "visualizable": 1,
+    "toolPrompt": "按照摘要描述查询凭证信息",
+    "toolType": "sql",
+    "datasourceId": "158",
+    "sqlTemplate": "WITH query_params AS (\n    SELECT \n        {year}::INT AS target_year,\n        '{digest}' AS digest_keyword,\n        '{company_names}' AS company_names_param,\n        '{months}' AS months_param\n),\ncompany_filters AS (\n    SELECT \n        TRIM(keyword) AS company_keyword\n    FROM (\n        SELECT unnest(string_to_array(nullif((SELECT company_names_param FROM query_params LIMIT 1), ''), ',')) AS keyword\n    ) t\n    WHERE keyword IS NOT NULL AND TRIM(keyword) <> ''\n),\nmonth_filters AS (\n    SELECT \n        month::INT AS target_month\n    FROM (\n        SELECT unnest(string_to_array(nullif((SELECT months_param FROM query_params LIMIT 1), ''), ',')) AS month\n    ) t\n    WHERE month IS NOT NULL AND TRIM(month) <> ''\n)\n\nSELECT \n    c.company_name AS 公司,\n    v.dbill_date AS 日期,\n    v.csign || LPAD(v.ino_id::text, 4, '0') AS 凭证号,\n    v.cdigest AS 摘要,\n    a.ccode AS 科目编码,\n    a.ccode_name AS 科目名称,\n    v.md AS 借方,\n    v.mc AS 贷方,\n    v.cbill AS 制单人\nFROM uf_voucher v\nJOIN uf_company c ON v.company_id = c.id\nJOIN uf_account_code a ON v.account_code_id = a.id\nCROSS JOIN query_params qp\nWHERE v.bdelete = false\n  AND v.iyear = qp.target_year\n  AND v.cdigest LIKE '%' || qp.digest_keyword || '%'\n  AND (\n      NOT EXISTS (SELECT 1 FROM company_filters)\n      OR EXISTS (\n          SELECT 1 FROM company_filters cf \n          WHERE c.company_name LIKE '%' || cf.company_keyword || '%'\n      )\n  )\n  AND (\n      NOT EXISTS (SELECT 1 FROM month_filters)\n      OR v.iperiod IN (SELECT target_month FROM month_filters)\n  )\nORDER BY c.company_name, v.dbill_date DESC, v.ino_id;",
+    "sqlParams": "[{\"type\":\"paragraph\",\"name\":\"digest\",\"displayName\":\"摘要关键字\",\"maxLength\":500,\"defaultValue\":\"\",\"required\":true},{\"type\":\"string\",\"name\":\"company_names\",\"displayName\":\"公司名称\",\"maxLength\":200,\"defaultValue\":\"\",\"required\":false},{\"type\":\"number\",\"name\":\"year\",\"displayName\":\"会计年度\",\"maxLength\":4,\"defaultValue\":\"2025\",\"required\":true},{\"type\":\"string\",\"name\":\"months\",\"displayName\":\"月份\",\"maxLength\":50,\"defaultValue\":\"\",\"required\":false}]",
+    "resultType": "list",
+    "sourceType": "ai",
+    "trainingTaskId": null,
+    "tableMetadataIds": "",
+    "executionCount": 0,
+    "visualizationConfigs": [],
+    "inputJsonSchema": "{}",
+    "outputJsonSchema": "{\"type\":\"object\",\"properties\":{\"data\":{\"type\":\"array\"}}}",
+    "lastExecutionTime": null
+  },
+  {
+    "id": "2000852304343240706",
+    "createBy": "bjt",
+    "createTime": "2025-12-16 16:55:38",
+    "updateBy": null,
+    "updateTime": null,
+    "serviceId": "2000821852159709186",
+    "uniqueName": "按科目查凭证",
+    "name": "ankemuchapingzheng_6a268ea1",
+    "description": "按照科目名称查询凭证信息",
+    "visualizable": 1,
+    "toolPrompt": "按照科目名称查询凭证信息",
+    "toolType": "sql",
+    "datasourceId": "158",
+    "sqlTemplate": "WITH query_params AS (\n    SELECT \n        {year}::INT AS target_year,\n        '{company_names}' AS company_names_param,\n        '{subject_name}' AS subject_keyword,\n        '{months}' AS months_param\n),\ncompany_filters AS (\n    SELECT \n        TRIM(keyword) AS company_keyword\n    FROM (\n        SELECT unnest(string_to_array(nullif((SELECT company_names_param FROM query_params LIMIT 1), ''), ',')) AS keyword\n    ) t\n    WHERE keyword IS NOT NULL AND TRIM(keyword) <> ''\n),\nmonth_filters AS (\n    SELECT \n        month::INT AS target_month\n    FROM (\n        SELECT unnest(string_to_array(nullif((SELECT months_param FROM query_params LIMIT 1), ''), ',')) AS month\n    ) t\n    WHERE month IS NOT NULL AND TRIM(month) <> ''\n),\n目标科目 AS (\n    SELECT DISTINCT a.id, a.ccode\n    FROM uf_account_code a\n    JOIN uf_company c ON a.company_id = c.id\n    CROSS JOIN query_params qp\n    WHERE a.ccode_name LIKE '%' || qp.subject_keyword || '%'\n      AND a.iyear = qp.target_year\n      AND (\n          NOT EXISTS (SELECT 1 FROM company_filters)\n          OR EXISTS (\n              SELECT 1 FROM company_filters cf \n              WHERE c.company_name LIKE '%' || cf.company_keyword || '%'\n          )\n      )\n),\n相关科目 AS (\n    SELECT a.id\n    FROM uf_account_code a\n    JOIN uf_company c ON a.company_id = c.id\n    CROSS JOIN query_params qp\n    WHERE a.iyear = qp.target_year\n      AND EXISTS (\n          SELECT 1 FROM 目标科目 t \n          WHERE a.ccode LIKE t.ccode || '%'\n      )\n      AND (\n          NOT EXISTS (SELECT 1 FROM company_filters)\n          OR EXISTS (\n              SELECT 1 FROM company_filters cf \n              WHERE c.company_name LIKE '%' || cf.company_keyword || '%'\n          )\n      )\n)\nSELECT \n    c.company_name AS 公司,\n    v.dbill_date AS 日期,\n    v.csign || LPAD(v.ino_id::text, 4, '0') AS 凭证号,\n    v.cdigest AS 摘要,\n    a.ccode AS 科目编码,\n    a.ccode_name AS 科目名称,\n    a.igrade AS 级次,\n    v.md AS 借方,\n    v.mc AS 贷方\nFROM uf_voucher v\nJOIN uf_company c ON v.company_id = c.id\nJOIN uf_account_code a ON v.account_code_id = a.id\nCROSS JOIN query_params qp\nWHERE v.bdelete = false\n  AND v.account_code_id IN (SELECT id FROM 相关科目)\n  AND v.iyear = qp.target_year\n  AND (\n      NOT EXISTS (SELECT 1 FROM company_filters)\n      OR EXISTS (\n          SELECT 1 FROM company_filters cf \n          WHERE c.company_name LIKE '%' || cf.company_keyword || '%'\n      )\n  )\n  AND (\n      NOT EXISTS (SELECT 1 FROM month_filters)\n      OR v.iperiod IN (SELECT target_month FROM month_filters)\n  )\nORDER BY c.company_name, a.ccode, v.dbill_date, v.ino_id;",
+    "sqlParams": "[{\"type\":\"select\",\"name\":\"subject_name\",\"displayName\":\"科目名称\",\"maxLength\":100,\"defaultValue\":\"\",\"required\":true,\"options\":[\"管理费用\",\"销售费用\",\"财务费用\",\"应收账款\",\"应付账款\"]},{\"type\":\"paragraph\",\"name\":\"company_names\",\"displayName\":\"公司名称列表\",\"maxLength\":300,\"defaultValue\":\"\",\"required\":false},{\"type\":\"number\",\"name\":\"year\",\"displayName\":\"年度\",\"maxLength\":4,\"defaultValue\":\"2025\",\"required\":true},{\"type\":\"select\",\"name\":\"months\",\"displayName\":\"期间月份\",\"maxLength\":30,\"defaultValue\":\"\",\"required\":false,\"options\":[\"1\",\"2\",\"3\",\"4\",\"5\",\"6\",\"7\",\"8\",\"9\",\"10\",\"11\",\"12\"]}]",
+    "resultType": "list",
+    "sourceType": "ai",
+    "trainingTaskId": null,
+    "tableMetadataIds": "",
+    "executionCount": 0,
+    "visualizationConfigs": [],
+    "inputJsonSchema": "{}",
+    "outputJsonSchema": "{\"type\":\"object\",\"properties\":{\"data\":{\"type\":\"array\"}}}",
+    "lastExecutionTime": null
+  },
+  {
+    "id": "2000852610904920066",
+    "createBy": "bjt",
+    "createTime": "2025-12-16 16:56:51",
+    "updateBy": null,
+    "updateTime": null,
+    "serviceId": "2000821852159709186",
+    "uniqueName": "按科目查余额",
+    "name": "ankemuchayue_68b302c2",
+    "description": "按照科目名称查询余额",
+    "visualizable": 1,
+    "toolPrompt": "按照科目名称查询余额",
+    "toolType": "sql",
+    "datasourceId": "158",
+    "sqlTemplate": "WITH query_params AS (\n    SELECT \n        {year}::INT AS target_year,\n        '{company_names}' AS company_names_param,\n        '{subject_name}' AS subject_keyword,\n        '{months}' AS months_param\n),\ncompany_filters AS (\n    SELECT \n        TRIM(keyword) AS company_keyword\n    FROM (\n        SELECT unnest(string_to_array(nullif((SELECT company_names_param FROM query_params LIMIT 1), ''), ',')) AS keyword\n    ) t\n    WHERE keyword IS NOT NULL AND TRIM(keyword) <> ''\n),\nmonth_filters AS (\n    SELECT \n        month::INT AS target_month\n    FROM (\n        SELECT unnest(string_to_array(nullif((SELECT months_param FROM query_params LIMIT 1), ''), ',')) AS month\n    ) t\n    WHERE month IS NOT NULL AND TRIM(month) <> ''\n),\n目标科目 AS (\n    SELECT DISTINCT a.id, a.ccode\n    FROM uf_account_code a\n    JOIN uf_company c ON a.company_id = c.id\n    CROSS JOIN query_params qp\n    WHERE a.ccode_name LIKE '%' || qp.subject_keyword || '%'\n      AND a.iyear = qp.target_year\n      AND (\n          NOT EXISTS (SELECT 1 FROM company_filters)\n          OR EXISTS (\n              SELECT 1 FROM company_filters cf \n              WHERE c.company_name LIKE '%' || cf.company_keyword || '%'\n          )\n      )\n),\n相关科目 AS (\n    SELECT a.id\n    FROM uf_account_code a\n    JOIN uf_company c ON a.company_id = c.id\n    CROSS JOIN query_params qp\n    WHERE a.iyear = qp.target_year\n      AND EXISTS (\n          SELECT 1 FROM 目标科目 t \n          WHERE a.ccode LIKE t.ccode || '%'\n      )\n      AND (\n          NOT EXISTS (SELECT 1 FROM company_filters)\n          OR EXISTS (\n              SELECT 1 FROM company_filters cf \n              WHERE c.company_name LIKE '%' || cf.company_keyword || '%'\n          )\n      )\n)\nSELECT \n    c.company_name AS 公司,\n    a.ccode AS 科目编码,\n    a.ccode_name AS 科目名称,\n    a.igrade AS 级次,\n    a.cclass AS 类别,\n    b.iyear AS 年,\n    b.iperiod AS 月,\n    b.cbegind_c AS 期初方向,\n    b.mb AS 期初余额,\n    b.md AS 借方发生,\n    b.mc AS 贷方发生,\n    b.cendd_c AS 期末方向,\n    b.me AS 期末余额\nFROM uf_account_balance b\nJOIN uf_company c ON b.company_id = c.id\nJOIN uf_account_code a ON b.account_code_id = a.id\nCROSS JOIN query_params qp\nWHERE b.iyear = qp.target_year\n  AND b.account_code_id IN (SELECT id FROM 相关科目)\n  AND (\n      NOT EXISTS (SELECT 1 FROM company_filters)\n      OR EXISTS (\n          SELECT 1 FROM company_filters cf \n          WHERE c.company_name LIKE '%' || cf.company_keyword || '%'\n      )\n  )\n  AND (\n      NOT EXISTS (SELECT 1 FROM month_filters)\n      OR b.iperiod IN (SELECT target_month FROM month_filters)\n  )\nORDER BY c.company_name, a.ccode, b.iperiod;",
+    "sqlParams": "[{\"type\":\"string\",\"name\":\"subject_name\",\"displayName\":\"科目名称关键字\",\"maxLength\":100,\"defaultValue\":\"银行存款\",\"required\":true},{\"type\":\"string\",\"name\":\"company_names\",\"displayName\":\"公司筛选\",\"maxLength\":150,\"defaultValue\":\"\",\"required\":false},{\"type\":\"number\",\"name\":\"year\",\"displayName\":\"查询年度\",\"maxLength\":4,\"defaultValue\":\"2025\",\"required\":true},{\"type\":\"number\",\"name\":\"months\",\"displayName\":\"查询月份\",\"maxLength\":2,\"defaultValue\":\"\",\"required\":false}]",
+    "resultType": "list",
+    "sourceType": "ai",
+    "trainingTaskId": null,
+    "tableMetadataIds": "",
+    "executionCount": 0,
+    "visualizationConfigs": [],
+    "inputJsonSchema": "{}",
+    "outputJsonSchema": "{\"type\":\"object\",\"properties\":{\"data\":{\"type\":\"array\"}}}",
+    "lastExecutionTime": null
+  }
+]
\ No newline at end of file
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/logs/.gitkeep b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/logs/.gitkeep
new file mode 100644
index 0000000..e69de29
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/main.py b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/main.py
new file mode 100644
index 0000000..dd4a0f6
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/main.py
@@ -0,0 +1,296 @@
+#!/usr/bin/env python3
+"""
+Workflow to MCP Server
+从 businessQueries.json 或 API 动态生成 MCP 工具
+
+支持两种模式:
+- local: 从本地 JSON 文件加载
+- api: 从远程 API 加载
+"""
+import json
+import os
+import logging
+import argparse
+import anyio
+
+import mcp.types as types
+from mcp.server import NotificationOptions, Server
+from mcp.server.models import InitializationOptions
+from mcp.server.stdio import stdio_server
+
+try:
+    from .schema_converter import convert_sql_params_to_input_schema
+    from .utils.api_client import execute_workflow, get_workflow_by_id
+    from .utils.env_config import get_workflow_id
+    from .utils.logger_config import setup_system_logging, get_logger
+except ImportError:
+    from schema_converter import convert_sql_params_to_input_schema
+    from utils.api_client import execute_workflow, get_workflow_by_id
+    from utils.env_config import get_workflow_id
+    from utils.logger_config import setup_system_logging, get_logger
+
+# 初始化日志系统
+setup_system_logging(app_name="lzwcai_workflow_to_mcp", log_level=logging.DEBUG)
+logger = get_logger(__name__)
+
+# 初始化 MCP Server
+server = Server("workflow_mcp_server")
+
+# 全局配置
+_tools_config: list[dict] = []
+_config: dict = {}
+
+
+def parse_arguments():
+    """解析命令行参数"""
+    parser = argparse.ArgumentParser(description="Workflow MCP Server")
+    parser.add_argument(
+        "--mode",
+        type=str,
+        choices=["local", "api"],
+        default="api",
+        help="数据加载模式: local(本地JSON) 或 api(远程API，默认)"
+    )
+    parser.add_argument(
+        "--json-path",
+        type=str,
+        default=None,
+        help="本地 JSON 文件路径 (local 模式)"
+    )
+    parser.add_argument(
+        "--workflow-id",
+        type=str,
+        default=None,
+        help="工作流ID (api 模式，可选，默认从环境变量 workflowId 获取)"
+    )
+    return parser.parse_args()
+
+
+class DataLoader:
+    """数据加载器基类"""
+    
+    def load(self) -> list[dict]:
+        raise NotImplementedError
+
+
+class LocalLoader(DataLoader):
+    """本地 JSON 文件加载器"""
+    
+    def __init__(self, json_path: str = None):
+        if json_path is None:
+            json_path = os.path.join(os.path.dirname(__file__), "businessQueries.json")
+        self.json_path = json_path
+    
+    def load(self) -> list[dict]:
+        try:
+            with open(self.json_path, "r", encoding="utf-8") as f:
+                data = json.load(f)
+                logger.info(f"从本地加载 {len(data)} 条配置: {self.json_path}")
+                return data
+        except FileNotFoundError:
+            logger.error(f"配置文件不存在: {self.json_path}")
+            return []
+        except json.JSONDecodeError as e:
+            logger.error(f"JSON 解析错误: {e}")
+            return []
+
+
+class ApiLoader(DataLoader):
+    """API 远程加载器 - 使用 get_workflow_by_id 获取工作流配置"""
+    
+    def __init__(self, workflow_id: str = None):
+        self.workflow_id = workflow_id or get_workflow_id()
+        logger.debug(f"ApiLoader 初始化，工作流ID: {self.workflow_id}")
+    
+    def load(self) -> list[dict]:
+        try:
+            if not self.workflow_id:
+                logger.error("未提供工作流ID，请设置环境变量 workflowId")
+                return []
+            
+            logger.info(f"开始从 API 加载工作流配置，工作流ID: {self.workflow_id}")
+            
+            # 调用 get_workflow_by_id 获取工作流配置
+            response = get_workflow_by_id(self.workflow_id)
+            
+            logger.debug(f"API 响应原始数据: {json.dumps(response, ensure_ascii=False, indent=2)}")
+            
+            # 返回格式: {code, data, msg}
+            if response.get("code") != 200:
+                logger.error(f"获取工作流配置失败: code={response.get('code')}, msg={response.get('msg')}")
+                return []
+            
+            data = response.get("data")
+            logger.debug(f"API 响应 data 字段: {json.dumps(data, ensure_ascii=False, indent=2) if data else 'None'}")
+            
+            # data 可能是单个对象或列表
+            if isinstance(data, dict):
+                result = [data]
+            elif isinstance(data, list):
+                result = data
+            else:
+                logger.warning(f"API 响应 data 字段类型异常: {type(data)}")
+                result = []
+            
+            logger.info(f"从 API 加载工作流配置成功，工作流ID: {self.workflow_id}, 配置数量: {len(result)}")
+            return result
+            
+        except Exception as e:
+            logger.error(f"API 请求失败: {e}", exc_info=True)
+            return []
+
+
+def create_loader(mode: str, **kwargs) -> DataLoader:
+    """
+    创建数据加载器
+    
+    Args:
+        mode: "local" 或 "api"
+        kwargs: 
+            - json_path: 本地 JSON 路径 (local 模式)
+            - workflow_id: 工作流ID (api 模式，可选，默认从环境变量获取)
+    """
+    if mode == "local":
+        return LocalLoader(json_path=kwargs.get("json_path"))
+    elif mode == "api":
+        return ApiLoader(workflow_id=kwargs.get("workflow_id"))
+    else:
+        raise ValueError(f"不支持的模式: {mode}")
+
+
+def init_tools(loader: DataLoader):
+    """初始化工具配置"""
+    global _tools_config
+    _tools_config = loader.load()
+    logger.info(f"已加载 {len(_tools_config)} 个工具配置")
+
+
+@server.list_tools()
+async def handle_list_tools() -> list[types.Tool]:
+    """列出所有可用工具"""
+    logger.info(f"收到 ListTools 请求，当前配置数量: {len(_tools_config)}")
+    tools = []
+    
+    for query in _tools_config:
+        name = query.get("name", "")
+        description = query.get("description") or query.get("toolPrompt") or query.get("uniqueName", "")
+        sql_params = query.get("sqlParams", "[]")
+        
+        logger.debug(f"处理工具配置: name={name}, description={description[:50] if description else 'None'}...")
+        
+        # 转换参数为 inputSchema
+        input_schema = convert_sql_params_to_input_schema(sql_params)
+        logger.debug(f"工具 {name} 的 inputSchema: {json.dumps(input_schema, ensure_ascii=False)}")
+        
+        tools.append(
+            types.Tool(
+                name=name,
+                description=description,
+                inputSchema=input_schema
+            )
+        )
+    
+    logger.info(f"ListTools 响应: 返回 {len(tools)} 个工具")
+    return tools
+
+
+@server.call_tool()
+async def handle_call_tool(
+    name: str, 
+    arguments: dict | None
+) -> list[types.TextContent | types.ImageContent | types.EmbeddedResource]:
+    """调用工具"""
+    logger.info(f"收到 CallTool 请求: name={name}, arguments={json.dumps(arguments, ensure_ascii=False) if arguments else 'None'}")
+    
+    # 查找对应的工具配置
+    tool_config = None
+    for query in _tools_config:
+        if query.get("name") == name:
+            tool_config = query
+            break
+    
+    if tool_config is None:
+        logger.error(f"未找到工具配置: {name}")
+        raise ValueError(f"未知工具: {name}")
+    
+    logger.debug(f"找到工具配置: {json.dumps(tool_config, ensure_ascii=False, indent=2)}")
+    
+    # 获取工作流ID
+    workflow_id = tool_config.get("workflowId") or get_workflow_id()
+    logger.info(f"使用工作流ID: {workflow_id}")
+    
+    # 构建请求数据
+    request_data = {
+        "workflowId": workflow_id,
+        "inputs": arguments or {}
+    }
+    
+    logger.info(f"执行工作流请求数据: {json.dumps(request_data, ensure_ascii=False, indent=2)}")
+    
+    try:
+        # 调用工作流执行API
+        result = execute_workflow(request_data)
+        logger.info(f"工作流执行成功: {workflow_id}")
+        logger.debug(f"工作流执行结果: {json.dumps(result, ensure_ascii=False, indent=2)}")
+    except Exception as e:
+        logger.error(f"工作流执行失败: {e}", exc_info=True)
+        result = {
+            "error": str(e),
+            "tool_name": name,
+            "workflow_id": workflow_id
+        }
+    
+    return [
+        types.TextContent(
+            type="text",
+            text=json.dumps(result, ensure_ascii=False, indent=2)
+        )
+    ]
+
+
+async def run_server():
+    """运行 MCP Server (stdio 模式)"""
+    async with stdio_server() as streams:
+        await server.run(
+            streams[0],
+            streams[1],
+            InitializationOptions(
+                server_name="workflow_mcp_server",
+                server_version="0.1.0",
+                capabilities=server.get_capabilities(
+                    notification_options=NotificationOptions(),
+                    experimental_capabilities={},
+                ),
+            ),
+        )
+
+
+def main():
+    """主入口"""
+    global _config
+    
+    logger.info("=" * 50)
+    logger.info("Workflow MCP Server 启动")
+    logger.info("=" * 50)
+    
+    # 解析命令行参数
+    args = parse_arguments()
+    _config = vars(args)
+    logger.info(f"命令行参数: {_config}")
+    
+    # 创建加载器并初始化工具
+    logger.info(f"使用模式: {args.mode}")
+    loader = create_loader(
+        mode=args.mode,
+        json_path=args.json_path,
+        workflow_id=args.workflow_id
+    )
+    init_tools(loader)
+    
+    logger.info("开始运行 MCP Server (stdio 模式)")
+    # 运行服务器
+    anyio.run(run_server)
+
+
+if __name__ == "__main__":
+    main()
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/pyproject.toml b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/pyproject.toml
new file mode 100644
index 0000000..1a485da
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/pyproject.toml
@@ -0,0 +1,18 @@
+[project]
+name = "lzwcai-workflow-to-mcp"
+version = "0.1.0"
+description = "从 businessQueries 动态生成 MCP 工具"
+readme = "README.md"
+requires-python = ">=3.10"
+dependencies = [
+    "mcp>=1.0.0",
+    "anyio>=4.0.0",
+    "requests>=2.28.0",
+]
+
+[project.scripts]
+workflow-mcp = "main:main"
+
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/schema_converter.py b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/schema_converter.py
new file mode 100644
index 0000000..405489d
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/schema_converter.py
@@ -0,0 +1,208 @@
+"""
+Schema 转换器
+将 sqlParams 数组格式转换为 MCP 工具需要的 JSON Schema 格式
+
+支持的类型:
+- string: 文本输入
+- paragraph: 段落/多行文本
+- select: 下拉选项
+- number: 数字输入
+"""
+import json
+from typing import Any
+
+
+def convert_param_to_schema_property(param: dict) -> tuple[str, dict, bool]:
+    """
+    将单个参数转换为 JSON Schema property
+    
+    Args:
+        param: 参数配置，格式如:
+            {
+                "type": "string",
+                "name": "company_names",
+                "displayName": "公司名称",
+                "maxLength": 200,
+                "defaultValue": "",
+                "required": true,
+                "options": ["选项1", "选项2"]  # 仅 select 类型
+            }
+    
+    Returns:
+        tuple: (property_name, property_schema, is_required)
+    """
+    param_type = param.get("type", "string")
+    param_name = param.get("name", "")
+    display_name = param.get("displayName", param_name)
+    default_value = param.get("defaultValue", "")
+    max_length = param.get("maxLength")
+    is_required = param.get("required", False)
+    options = param.get("options", [])
+    
+    property_schema = {
+        "description": display_name
+    }
+    
+    if param_type == "string":
+        property_schema["type"] = "string"
+        if max_length:
+            property_schema["maxLength"] = max_length
+            
+    elif param_type == "paragraph":
+        property_schema["type"] = "string"
+        property_schema["format"] = "paragraph"
+        if max_length:
+            property_schema["maxLength"] = max_length
+            
+    elif param_type == "select":
+        property_schema["type"] = "string"
+        if options:
+            property_schema["enum"] = options
+            
+    elif param_type == "number":
+        property_schema["type"] = "number"
+        
+    else:
+        # 默认当作 string 处理
+        property_schema["type"] = "string"
+    
+    # 添加默认值
+    if default_value not in (None, ""):
+        if param_type == "number":
+            try:
+                property_schema["default"] = int(default_value) if str(default_value).isdigit() else float(default_value)
+            except (ValueError, TypeError):
+                property_schema["default"] = default_value
+        else:
+            property_schema["default"] = default_value
+    
+    return param_name, property_schema, is_required
+
+
+def convert_sql_params_to_input_schema(sql_params: str | list) -> dict:
+    """
+    将 sqlParams 转换为 MCP 工具的 inputSchema
+    
+    Args:
+        sql_params: sqlParams 字段值，可以是 JSON 字符串或已解析的列表
+            格式: [{"type": "string", "name": "xxx", ...}, ...]
+    
+    Returns:
+        dict: MCP 工具的 inputSchema，格式如:
+            {
+                "type": "object",
+                "properties": {...},
+                "required": [...]
+            }
+    """
+    # 解析 JSON 字符串
+    if isinstance(sql_params, str):
+        try:
+            params_list = json.loads(sql_params)
+        except json.JSONDecodeError:
+            return {"type": "object", "properties": {}, "required": []}
+    else:
+        params_list = sql_params
+    
+    if not isinstance(params_list, list):
+        return {"type": "object", "properties": {}, "required": []}
+    
+    input_schema = {
+        "type": "object",
+        "properties": {},
+        "required": []
+    }
+    
+    for param in params_list:
+        if not isinstance(param, dict):
+            continue
+            
+        name, schema, is_required = convert_param_to_schema_property(param)
+        
+        if name:
+            input_schema["properties"][name] = schema
+            if is_required:
+                input_schema["required"].append(name)
+    
+    return input_schema
+
+
+def convert_business_query_to_tool(query: dict) -> dict:
+    """
+    将单个 businessQuery 转换为 MCP Tool 配置
+    
+    Args:
+        query: businessQuery 对象
+    
+    Returns:
+        dict: MCP Tool 配置
+            {
+                "name": "工具名称",
+                "description": "工具描述",
+                "inputSchema": {...}
+            }
+    """
+    name = query.get("name", "")
+    description = query.get("description") or query.get("toolPrompt") or query.get("uniqueName", "")
+    sql_params = query.get("sqlParams", "[]")
+    
+    input_schema = convert_sql_params_to_input_schema(sql_params)
+    
+    return {
+        "name": name,
+        "description": description,
+        "inputSchema": input_schema,
+        # 保留原始数据供后续使用
+        "_raw": {
+            "id": query.get("id"),
+            "uniqueName": query.get("uniqueName"),
+            "sqlTemplate": query.get("sqlTemplate"),
+            "datasourceId": query.get("datasourceId"),
+            "toolType": query.get("toolType"),
+        }
+    }
+
+
+def convert_all_queries_to_tools(queries: list[dict]) -> list[dict]:
+    """
+    将所有 businessQueries 转换为 MCP Tools 列表
+    
+    Args:
+        queries: businessQueries 列表
+    
+    Returns:
+        list: MCP Tools 配置列表
+    """
+    tools = []
+    for query in queries:
+        tool = convert_business_query_to_tool(query)
+        if tool["name"]:
+            tools.append(tool)
+    return tools
+
+
+# 测试用
+if __name__ == "__main__":
+    # 测试单个参数转换
+    test_param = {
+        "type": "select",
+        "name": "subject_name",
+        "displayName": "科目名称",
+        "maxLength": 100,
+        "defaultValue": "",
+        "required": True,
+        "options": ["管理费用", "销售费用", "财务费用"]
+    }
+    
+    name, schema, required = convert_param_to_schema_property(test_param)
+    print(f"参数名: {name}")
+    print(f"Schema: {json.dumps(schema, ensure_ascii=False, indent=2)}")
+    print(f"必填: {required}")
+    print()
+    
+    # 测试完整 sqlParams 转换
+    test_sql_params = '[{"type":"paragraph","name":"digest","displayName":"摘要关键字","maxLength":500,"defaultValue":"","required":true},{"type":"string","name":"company_names","displayName":"公司名称","maxLength":200,"defaultValue":"","required":false},{"type":"number","name":"year","displayName":"会计年度","maxLength":4,"defaultValue":"2025","required":true}]'
+    
+    input_schema = convert_sql_params_to_input_schema(test_sql_params)
+    print("InputSchema:")
+    print(json.dumps(input_schema, ensure_ascii=False, indent=2))
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__init__.py b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__init__.py
new file mode 100644
index 0000000..53da55d
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__init__.py
@@ -0,0 +1,33 @@
+"""Utils package for lzwcai_workflow_to_mcp"""
+
+from .json_helper import load_json
+from .name_helper import generate_tool_name
+from .schema_helper import generate_input_schema, validate_input_schema
+from .api_client import (
+    WorkflowAPIClient, 
+    get_workflow_by_id, 
+    execute_workflow, 
+    process_workflow_response,
+    get_default_client,
+    DEFAULT_TIMEOUT,
+    WORKFLOW_EXECUTE_TIMEOUT
+)
+from .env_config import get_workflow_id, get_backend_base_url, get_env_config, set_env_variable
+
+__all__ = [
+    'load_json', 
+    'generate_tool_name', 
+    'generate_input_schema', 
+    'validate_input_schema',
+    'WorkflowAPIClient',
+    'get_workflow_by_id',
+    'execute_workflow',
+    'process_workflow_response',
+    'get_default_client',
+    'DEFAULT_TIMEOUT',
+    'WORKFLOW_EXECUTE_TIMEOUT',
+    'get_workflow_id',
+    'get_backend_base_url',
+    'get_env_config',
+    'set_env_variable'
+]
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/__init__.cpython-312.pyc b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/__init__.cpython-312.pyc
new file mode 100644
index 0000000..a94cf01
Binary files /dev/null and b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/__init__.cpython-312.pyc differ
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/api_client.cpython-312.pyc b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/api_client.cpython-312.pyc
new file mode 100644
index 0000000..87ce3b9
Binary files /dev/null and b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/api_client.cpython-312.pyc differ
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/env_config.cpython-312.pyc b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/env_config.cpython-312.pyc
new file mode 100644
index 0000000..0c048b2
Binary files /dev/null and b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/env_config.cpython-312.pyc differ
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/json_helper.cpython-312.pyc b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/json_helper.cpython-312.pyc
new file mode 100644
index 0000000..8aad326
Binary files /dev/null and b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/json_helper.cpython-312.pyc differ
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/logger_config.cpython-312.pyc b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/logger_config.cpython-312.pyc
new file mode 100644
index 0000000..1da07dc
Binary files /dev/null and b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/logger_config.cpython-312.pyc differ
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/name_helper.cpython-312.pyc b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/name_helper.cpython-312.pyc
new file mode 100644
index 0000000..b61c36d
Binary files /dev/null and b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/name_helper.cpython-312.pyc differ
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/schema_helper.cpython-312.pyc b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/schema_helper.cpython-312.pyc
new file mode 100644
index 0000000..9eb072d
Binary files /dev/null and b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/__pycache__/schema_helper.cpython-312.pyc differ
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/api_client.py b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/api_client.py
new file mode 100644
index 0000000..ef72edb
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/api_client.py
@@ -0,0 +1,307 @@
+"""
+工作流API调用客户端
+用于调用工作流管理和执行接口
+"""
+
+import httpx
+import json
+from typing import Dict, Any, Optional, List
+
+try:
+    from .env_config import get_backend_base_url, get_workflow_execute_key
+    from .logger_config import get_logger
+except ImportError:
+    from env_config import get_backend_base_url, get_workflow_execute_key
+    from logger_config import get_logger
+
+logger = get_logger(__name__)
+
+# 默认超时配置（秒）
+DEFAULT_TIMEOUT = 30.0  # 普通请求超时
+WORKFLOW_EXECUTE_TIMEOUT = 300.0  # 工作流执行超时（5分钟）
+
+
+class WorkflowAPIClient:
+    """工作流API客户端"""
+    
+    def __init__(
+        self, 
+        base_url: Optional[str] = None,
+        token: Optional[str] = None,
+        default_timeout: float = DEFAULT_TIMEOUT,
+        execute_timeout: float = WORKFLOW_EXECUTE_TIMEOUT
+    ):
+        """
+        初始化API客户端
+        
+        Args:
+            base_url: API基础URL（默认从环境变量 backendBaseUrl 读取）
+            token: 认证令牌（默认从环境变量 workflowExecuteKey 读取）
+            default_timeout: 普通请求超时时间（秒），默认30秒
+            execute_timeout: 工作流执行超时时间（秒），默认300秒（5分钟）
+        """
+        if base_url is None:
+            base_url = get_backend_base_url()
+        
+        if token is None:
+            token = get_workflow_execute_key()
+        
+        self.base_url = base_url.rstrip('/')
+        self.token = token
+        self.default_timeout = default_timeout
+        self.execute_timeout = execute_timeout
+        self._client: Optional[httpx.Client] = None
+        
+        logger.debug(f"WorkflowAPIClient 初始化: base_url={self.base_url}, token={'*' * 10 if self.token else 'None'}, execute_timeout={self.execute_timeout}s")
+    
+    @property
+    def client(self) -> httpx.Client:
+        """懒加载 HTTP 客户端"""
+        if self._client is None:
+            self._client = httpx.Client(timeout=self.default_timeout)
+        return self._client
+        
+    def _get_headers(self) -> Dict[str, str]:
+        """获取请求头"""
+        return {
+            'X-API-Key': f'Bearer {self.token}',
+        }
+    
+    def get_workflow_by_id(self, workflow_id: str) -> Dict[str, Any]:
+        """
+        根据工作流ID获取工作流信息
+        
+        Args:
+            workflow_id: 工作流ID
+            
+        Returns:
+            API响应数据
+            
+        Raises:
+            Exception: 请求失败时抛出
+        """
+        url = f"{self.base_url}/system/workflowManage/getByWorkflowId/{workflow_id}"
+        
+        try:
+            logger.info(f"[API请求] GET {url}")
+            logger.debug(f"[API请求] Headers: {self._get_headers()}")
+            
+            response = self.client.get(
+                url,
+                headers=self._get_headers()
+            )
+            
+            logger.info(f"[API响应] HTTP {response.status_code}")
+            logger.debug(f"[API响应] Headers: {dict(response.headers)}")
+            
+            response.raise_for_status()
+            data = response.json()
+            
+            logger.info(f"[API响应] 获取工作流配置成功: workflow_id={workflow_id}")
+            logger.debug(f"[API响应] Body: {json.dumps(data, ensure_ascii=False, indent=2)}")
+            
+            return data
+            
+        except httpx.TimeoutException:
+            error_msg = f"API请求超时: {url}"
+            logger.error(f"[API错误] {error_msg}")
+            raise Exception(error_msg)
+
+        except httpx.HTTPStatusError as e:
+            error_msg = f"API请求失败 (HTTP {e.response.status_code}): {url}"
+            logger.error(f"[API错误] {error_msg}")
+            logger.error(f"[API错误] 响应内容: {e.response.text}")
+            raise Exception(error_msg)
+            
+        except httpx.RequestError as e:
+            error_msg = f"API请求异常: {url}, 错误: {str(e)}"
+            logger.error(f"[API错误] {error_msg}")
+            raise Exception(error_msg)
+            
+        except Exception as e:
+            error_msg = f"处理API响应时出错: {str(e)}"
+            logger.error(f"[API错误] {error_msg}", exc_info=True)
+            raise Exception(error_msg)
+    
+    def execute_workflow(self, request_data: Dict[str, Any]) -> Dict[str, Any]:
+        """
+        执行工作流
+        
+        Args:
+            request_data: 请求数据，包含工作流执行所需的参数
+                
+        Returns:
+            API响应数据
+            
+        Raises:
+            Exception: 请求失败时抛出
+        """
+        url = f"{self.base_url}/open/workflow/execute"
+        
+        try:
+            headers = self._get_headers()
+            headers['Content-Type'] = 'application/json'
+            headers['Accept'] = '*/*'
+            
+            logger.info(f"[API请求] POST {url} (超时: {self.execute_timeout}s)")
+            logger.debug(f"[API请求] Headers: {headers}")
+            logger.debug(f"[API请求] Body: {json.dumps(request_data, ensure_ascii=False, indent=2)}")
+            
+            # 使用更长的超时时间执行工作流
+            response = self.client.post(
+                url,
+                headers=headers,
+                json=request_data,
+                timeout=self.execute_timeout  # 使用工作流执行专用超时时间
+            )
+            
+            logger.info(f"[API响应] HTTP {response.status_code}")
+            logger.debug(f"[API响应] Headers: {dict(response.headers)}")
+            
+            response.raise_for_status()
+            data = response.json()
+            
+            logger.info(f"[API响应] 执行工作流成功")
+            logger.debug(f"[API响应] Body: {json.dumps(data, ensure_ascii=False, indent=2)}")
+            
+            return data
+            
+        except httpx.TimeoutException:
+            error_msg = f"执行工作流API请求超时: {url}"
+            logger.error(f"[API错误] {error_msg}")
+            raise Exception(error_msg)
+            
+        except httpx.HTTPStatusError as e:
+            error_msg = f"执行工作流API请求失败 (HTTP {e.response.status_code}): {url}"
+            logger.error(f"[API错误] {error_msg}")
+            logger.error(f"[API错误] 响应内容: {e.response.text}")
+            raise Exception(error_msg)
+            
+        except httpx.RequestError as e:
+            error_msg = f"执行工作流API请求异常: {url}, 错误: {str(e)}"
+            logger.error(f"[API错误] {error_msg}")
+            raise Exception(error_msg)
+            
+        except Exception as e:
+            error_msg = f"处理执行工作流API响应时出错: {str(e)}"
+            logger.error(f"[API错误] {error_msg}", exc_info=True)
+            raise Exception(error_msg)
+    
+    def close(self):
+        """关闭HTTP客户端"""
+        if self._client is not None:
+            self._client.close()
+            self._client = None
+    
+    def __enter__(self):
+        """支持 context manager"""
+        return self
+    
+    def __exit__(self, exc_type, exc_val, exc_tb):
+        """退出时关闭客户端"""
+        self.close()
+        return False
+
+
+# 懒加载的默认客户端
+_default_client: Optional[WorkflowAPIClient] = None
+
+
+def get_default_client() -> WorkflowAPIClient:
+    """获取默认客户端（懒加载）"""
+    global _default_client
+    if _default_client is None:
+        _default_client = WorkflowAPIClient()
+    return _default_client
+
+
+def get_workflow_by_id(workflow_id: str, base_url: Optional[str] = None, token: Optional[str] = None) -> Dict[str, Any]:
+    """
+    便捷函数：根据工作流ID获取工作流信息
+    
+    Args:
+        workflow_id: 工作流ID
+        base_url: API基础URL（可选）
+        token: 认证令牌（可选）
+        
+    Returns:
+        API响应数据
+    """
+    if base_url or token:
+        with WorkflowAPIClient(base_url=base_url, token=token) as client:
+            return client.get_workflow_by_id(workflow_id)
+    else:
+        return get_default_client().get_workflow_by_id(workflow_id)
+
+
+def execute_workflow(
+    request_data: Dict[str, Any], 
+    base_url: Optional[str] = None, 
+    token: Optional[str] = None,
+    timeout: float = WORKFLOW_EXECUTE_TIMEOUT
+) -> Dict[str, Any]:
+    """
+    便捷函数：执行工作流
+    
+    Args:
+        request_data: 请求数据
+        base_url: API基础URL（可选）
+        token: 认证令牌（可选）
+        timeout: 超时时间（秒），默认300秒（5分钟）
+        
+    Returns:
+        API响应数据
+    """
+    if base_url or token:
+        with WorkflowAPIClient(base_url=base_url, token=token, execute_timeout=timeout) as client:
+            return client.execute_workflow(request_data)
+    else:
+        # 更新默认客户端的超时时间
+        default_client = get_default_client()
+        default_client.execute_timeout = timeout
+        return default_client.execute_workflow(request_data)
+
+
+def process_workflow_response(response: Dict[str, Any]) -> List[Dict[str, Any]]:
+    """
+    处理API响应数据，映射为工作流配置格式
+    
+    Args:
+        response: API原始响应数据
+        
+    Returns:
+        处理后的工作流配置列表
+    """
+    try:
+        data_list = response.get("data", [])
+        
+        # 如果 data 是单个对象而不是列表，转换为列表
+        if isinstance(data_list, dict):
+            data_list = [data_list]
+        
+        workflows = []
+        for workflow in data_list:
+            # 解析 inputParams 字符串为 JSON 对象
+            input_params = workflow.get("inputParams", {})
+            if isinstance(input_params, str):
+                try:
+                    input_params = json.loads(input_params)
+                except json.JSONDecodeError:
+                    input_params = {}
+            
+            # 映射字段
+            config = {
+                "id": workflow.get("id"),
+                "workflowId": workflow.get("workflowId"),
+                "workflowName": workflow.get("workflowName") or workflow.get("name"),
+                "workflowDescription": workflow.get("workflowDescription") or workflow.get("description"),
+                "inputParams": input_params
+            }
+            workflows.append(config)
+        
+        logger.info(f"成功处理 {len(workflows)} 条工作流数据")
+        return workflows
+        
+    except Exception as e:
+        logger.error(f"处理API响应数据失败: {e}", exc_info=True)
+        raise
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/env_config.py b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/env_config.py
new file mode 100644
index 0000000..eac8f35
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/env_config.py
@@ -0,0 +1,85 @@
+"""环境变量配置模块"""
+
+import os
+from typing import Optional
+
+
+def get_workflow_id(default: str = "") -> str:
+    """
+    获取工作流ID环境变量
+    
+    Args:
+        default: 默认值（默认为 ""）
+        
+    Returns:
+        str: 工作流ID
+        
+    Environment Variables:
+        workflowId: 工作流ID
+    """
+    return os.environ.get("workflowId", default)
+
+
+def get_backend_base_url(default: str = "http://lzwcai-demp-corp-manager:8086") -> str:
+    """
+    获取后端API基础URL环境变量
+    
+    Args:
+        default: 默认值（默认为 "http://lzwcai-demp-corp-manager:8086"）
+        
+    Returns:
+        str: 后端API基础URL
+        
+    Environment Variables:
+        backendBaseUrl: 后端API基础URL
+    """
+    return os.environ.get("backendBaseUrl", default)
+
+
+def get_workflow_execute_key(default: str = "") -> str:
+    """
+    获取工作流执行密钥（Token）环境变量
+    
+    Args:
+        default: 默认值（默认为 ""）
+        
+    Returns:
+        str: 工作流执行密钥
+        
+    Environment Variables:
+        workflowExecuteKey: 工作流执行密钥/Token
+    """
+    return os.environ.get("workflowExecuteKey", default)
+
+
+def get_env_config() -> dict:
+    """
+    获取所有环境配置
+    
+    Returns:
+        dict: 包含所有配置的字典
+        
+    Example:
+        config = get_env_config()
+        print(config['workflow_id'])       # 输出: ""
+        print(config['backend_base_url'])  # 输出: "http://lzwcai-demp-corp-manager:8086"
+    """
+    return {
+        "workflow_id": get_workflow_id(),
+        "backend_base_url": get_backend_base_url(),
+        "workflow_execute_key": get_workflow_execute_key()
+    }
+
+
+def set_env_variable(key: str, value: str) -> None:
+    """
+    设置环境变量（仅在当前进程中有效）
+    
+    Args:
+        key: 环境变量名
+        value: 环境变量值
+        
+    Example:
+        set_env_variable("workflowId", "1234567890")
+    """
+    os.environ[key] = value
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/json_helper.py b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/json_helper.py
new file mode 100644
index 0000000..cd6894e
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/json_helper.py
@@ -0,0 +1,59 @@
+"""JSON 文件读取工具"""
+
+import json
+from pathlib import Path
+from typing import Any, Union
+
+
+def load_json(json_path: Union[str, Path]) -> Any:
+    """
+    读取 JSON 文件并返回其内容
+    
+    Args:
+        json_path: JSON 文件的路径（支持字符串或 Path 对象）
+        
+    Returns:
+        JSON 文件中解析后的数据（可以是字典、列表或其他 JSON 类型）
+        
+    Raises:
+        FileNotFoundError: 当文件不存在时
+        json.JSONDecodeError: 当 JSON 格式无效时
+        Exception: 其他读取错误
+        
+    Example:
+        >>> data = load_json('config.json')
+        >>> print(data)
+        {'key': 'value'}
+        
+        >>> data = load_json(Path('data/users.json'))
+        >>> print(data)
+        [{'id': 1, 'name': 'Alice'}]
+    """
+    try:
+        # 转换为 Path 对象
+        path = Path(json_path)
+        
+        # 检查文件是否存在
+        if not path.exists():
+            raise FileNotFoundError(f"JSON 文件不存在: {json_path}")
+        
+        # 检查是否为文件
+        if not path.is_file():
+            raise ValueError(f"路径不是一个文件: {json_path}")
+        
+        # 读取并解析 JSON 文件
+        with open(path, 'r', encoding='utf-8') as f:
+            data = json.load(f)
+        
+        return data
+        
+    except json.JSONDecodeError as e:
+        raise json.JSONDecodeError(
+            f"JSON 格式错误: {e.msg}",
+            e.doc,
+            e.pos
+        )
+    except FileNotFoundError:
+        raise
+    except Exception as e:
+        raise Exception(f"读取 JSON 文件时发生错误: {str(e)}")
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/logger_config.py b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/logger_config.py
new file mode 100644
index 0000000..38efa16
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/logger_config.py
@@ -0,0 +1,174 @@
+# -*- coding: utf-8 -*-
+"""
+统一日志配置模块
+提供系统级别的日志配置和管理
+"""
+
+import os
+import sys
+import logging
+from logging.handlers import RotatingFileHandler, TimedRotatingFileHandler
+from datetime import datetime
+from pathlib import Path
+
+
+class LoggerConfig:
+    """日志配置管理类"""
+    
+    def __init__(self, logs_dir: str = None):
+        """初始化日志配置
+        
+        Args:
+            logs_dir: 日志目录路径，默认为项目根目录下的logs文件夹
+        """
+        if logs_dir:
+            self.logs_dir = Path(logs_dir)
+        else:
+            project_root = Path(__file__).parent.parent
+            self.logs_dir = project_root / "logs"
+        
+        self.logs_dir.mkdir(exist_ok=True)
+        
+        self.log_format = '%(asctime)s - %(name)s - %(levelname)s - [%(filename)s:%(lineno)d] - %(message)s'
+        self.date_format = '%Y-%m-%d %H:%M:%S'
+        self.log_level = self._get_log_level_from_env()
+        self._initialized = False
+    
+    def _get_log_level_from_env(self) -> int:
+        log_level_str = os.getenv('LOG_LEVEL', 'INFO').upper()
+        level_mapping = {
+            'DEBUG': logging.DEBUG,
+            'INFO': logging.INFO,
+            'WARNING': logging.WARNING,
+            'WARN': logging.WARNING,
+            'ERROR': logging.ERROR,
+            'CRITICAL': logging.CRITICAL,
+            'FATAL': logging.CRITICAL
+        }
+        return level_mapping.get(log_level_str, logging.INFO)
+    
+    def setup_logging(self, 
+                     app_name: str = "lzwcai_workflow_to_mcp",
+                     log_level: int = logging.INFO,
+                     max_file_size: int = 10 * 1024 * 1024,
+                     backup_count: int = 5,
+                     console_output: bool = True) -> logging.Logger:
+        if self._initialized:
+            return logging.getLogger()
+        
+        root_logger = logging.getLogger()
+        root_logger.setLevel(log_level)
+        
+        for handler in root_logger.handlers[:]:
+            root_logger.removeHandler(handler)
+        
+        formatter = logging.Formatter(self.log_format, self.date_format)
+
+        # 1. 主日志文件 - 按大小滚动
+        main_log_file = self.logs_dir / f"{app_name}.log"
+        file_handler = RotatingFileHandler(
+            main_log_file,
+            maxBytes=max_file_size,
+            backupCount=backup_count,
+            encoding='utf-8'
+        )
+        file_handler.setLevel(log_level)
+        file_handler.setFormatter(formatter)
+        root_logger.addHandler(file_handler)
+        
+        # 2. 错误日志文件
+        error_log_file = self.logs_dir / f"{app_name}_error.log"
+        error_handler = RotatingFileHandler(
+            error_log_file,
+            maxBytes=max_file_size,
+            backupCount=backup_count,
+            encoding='utf-8'
+        )
+        error_handler.setLevel(logging.ERROR)
+        error_handler.setFormatter(formatter)
+        root_logger.addHandler(error_handler)
+        
+        # 3. 按日期滚动的日志文件
+        daily_log_file = self.logs_dir / f"{app_name}_daily.log"
+        daily_handler = TimedRotatingFileHandler(
+            daily_log_file,
+            when='midnight',
+            interval=1,
+            backupCount=30,
+            encoding='utf-8'
+        )
+        daily_handler.setLevel(log_level)
+        daily_handler.setFormatter(formatter)
+        daily_handler.suffix = "%Y-%m-%d"
+        root_logger.addHandler(daily_handler)
+        
+        # 4. 控制台输出 (MCP协议使用stdio时，必须将日志输出到stderr)
+        if console_output:
+            console_handler = logging.StreamHandler(sys.stderr)
+            console_handler.setLevel(log_level)
+            console_formatter = logging.Formatter(
+                '%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+                self.date_format
+            )
+            console_handler.setFormatter(console_formatter)
+            root_logger.addHandler(console_handler)
+        
+        self._initialized = True
+        root_logger.info(f"日志系统初始化完成 - 日志目录: {self.logs_dir}")
+        
+        return root_logger
+    
+    def get_module_logger(self, module_name: str) -> logging.Logger:
+        return logging.getLogger(module_name)
+    
+    def create_component_logger(self, 
+                               component_name: str,
+                               log_file: str = None,
+                               log_level: int = None) -> logging.Logger:
+        logger = logging.getLogger(component_name)
+        
+        if log_file:
+            component_log_file = self.logs_dir / log_file
+            handler = RotatingFileHandler(
+                component_log_file,
+                maxBytes=5 * 1024 * 1024,
+                backupCount=3,
+                encoding='utf-8'
+            )
+            
+            formatter = logging.Formatter(self.log_format, self.date_format)
+            handler.setFormatter(formatter)
+            
+            if log_level:
+                handler.setLevel(log_level)
+            
+            logger.addHandler(handler)
+        
+        return logger
+    
+    def setup_mcp_logging(self) -> logging.Logger:
+        return self.create_component_logger(
+            "mcp_services",
+            "mcp_services.log",
+            logging.DEBUG
+        )
+    
+    def setup_api_logging(self) -> logging.Logger:
+        return self.create_component_logger(
+            "api_requests",
+            "api_requests.log",
+            logging.INFO
+        )
+
+
+# 全局日志配置实例
+logger_config = LoggerConfig()
+
+
+def setup_system_logging(app_name: str = "lzwcai_workflow_to_mcp", 
+                        log_level: int = logging.INFO) -> logging.Logger:
+    return logger_config.setup_logging(app_name, log_level)
+
+
+def get_logger(name: str) -> logging.Logger:
+    return logger_config.get_module_logger(name)
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/name_helper.py b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/name_helper.py
new file mode 100644
index 0000000..7a5499d
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/name_helper.py
@@ -0,0 +1,40 @@
+# -*- coding: utf-8 -*-
+"""
+名称生成工具模块
+"""
+
+from pypinyin import lazy_pinyin, Style
+import logging
+
+logger = logging.getLogger(__name__)
+
+
+def generate_tool_name(business_name: str, tool_id: str) -> str:
+    """
+    根据业务名称和ID生成工具名称
+    格式: tool_拼音_id
+    
+    Args:
+        business_name: 业务名称（中文）
+        tool_id: 工具ID
+        
+    Returns:
+        str: 格式化的工具名称
+    """
+    try:
+        # 将中文转换为拼音（无音调，小写）
+        pinyin_list = lazy_pinyin(business_name, style=Style.NORMAL)
+        # 拼接拼音
+        pinyin_str = ''.join(pinyin_list)
+        
+        # 将 ID 中的 '-' 替换为 '_'
+        formatted_id = tool_id.replace('-', '_')
+        
+        # 组合成最终的工具名称
+        tool_name = f"workflow_{pinyin_str}_{formatted_id}"
+        
+        return tool_name
+    except Exception as e:
+        logger.error(f"生成工具名称失败: {business_name}, {tool_id}, 错误: {e}", exc_info=True)
+        # 降级处理：如果拼音转换失败，使用 ID
+        return f"workflow_{tool_id.replace('-', '_')}"
diff --git a/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/schema_helper.py b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/schema_helper.py
new file mode 100644
index 0000000..ee92867
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/lzwcai_workflow_to_mcp/utils/schema_helper.py
@@ -0,0 +1,94 @@
+# -*- coding: utf-8 -*-
+"""
+Schema 生成工具模块
+"""
+
+from typing import Any, Dict, List
+
+
+def generate_input_schema(parameters: Dict[str, Any]) -> Dict[str, Any]:
+    """
+    从查询配置的参数定义生成 MCP 工具的 inputSchema
+    
+    此函数会保留完整的 JSON Schema 信息，包括：
+    - type: Schema 类型（通常是 "object"）
+    - required: 必填字段列表
+    - properties: 属性定义（包括每个属性的 type, description, format, examples 等）
+    - description: Schema 的整体描述（如果有）
+    - 以及其他任何 JSON Schema 标准字段
+    
+    Args:
+        parameters: 查询配置中的参数定义字典，应该是一个完整的 JSON Schema 对象
+        
+    Returns:
+        Dict[str, Any]: 符合 JSON Schema 规范的 inputSchema 对象
+    """
+    # 如果 parameters 本身就是一个完整的 JSON Schema 对象，直接使用
+    # 但确保至少包含 type 和 properties
+    if not parameters:
+        # 如果 parameters 为空，返回一个空的 object schema
+        return {
+            "type": "object",
+            "properties": {},
+            "required": []
+        }
+    
+    # 深拷贝 parameters 以避免修改原始数据
+    input_schema = dict(parameters)
+    
+    # 确保必需的字段存在
+    if "type" not in input_schema:
+        input_schema["type"] = "object"
+    
+    if "properties" not in input_schema:
+        input_schema["properties"] = {}
+    
+    if "required" not in input_schema:
+        input_schema["required"] = []
+    
+    return input_schema
+
+
+def validate_input_schema(schema: Dict[str, Any]) -> tuple[bool, str]:
+    """
+    验证 inputSchema 是否符合基本的 JSON Schema 规范
+    
+    Args:
+        schema: 要验证的 schema 对象
+        
+    Returns:
+        tuple[bool, str]: (是否有效, 错误消息或成功消息)
+    """
+    if not isinstance(schema, dict):
+        return False, "Schema 必须是一个字典对象"
+    
+    if schema.get("type") != "object":
+        return False, "Schema 的 type 字段必须是 'object'"
+    
+    if "properties" not in schema:
+        return False, "Schema 必须包含 properties 字段"
+    
+    if not isinstance(schema.get("properties"), dict):
+        return False, "Schema 的 properties 字段必须是一个字典对象"
+    
+    # 验证 required 字段（如果存在）
+    if "required" in schema:
+        required = schema["required"]
+        if not isinstance(required, list):
+            return False, "Schema 的 required 字段必须是一个列表"
+        
+        # 验证所有 required 的字段都在 properties 中定义
+        properties = schema["properties"]
+        for field in required:
+            if field not in properties:
+                return False, f"必填字段 '{field}' 未在 properties 中定义"
+    
+    # 验证 properties 中每个字段的定义
+    for prop_name, prop_def in schema["properties"].items():
+        if not isinstance(prop_def, dict):
+            return False, f"属性 '{prop_name}' 的定义必须是一个字典对象"
+        
+        if "type" not in prop_def:
+            return False, f"属性 '{prop_name}' 必须包含 type 字段"
+    
+    return True, "Schema 验证通过"
diff --git a/lzwcai_workflow_to_mcp/main.py b/lzwcai_workflow_to_mcp/main.py
new file mode 100644
index 0000000..80107e5
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/main.py
@@ -0,0 +1,16 @@
+"""
+Entry point for lzwcai-workflow-to-mcp
+Runs the MCP server for workflow execution
+"""
+
+import os
+
+if __name__ == "__main__":
+    # 设置环境变量
+    os.environ["workflowId"] = "2002300699510763521"
+    os.environ["workflowExecuteKey"] = "wf_buh230o9iaea4n6aefsddcexa7p27ydl"
+    os.environ["backendBaseUrl"] = "http://192.168.2.236:8088"
+    
+    # Import and run the actual MCP server
+    from lzwcai_workflow_to_mcp.main import main
+    main()
diff --git a/lzwcai_workflow_to_mcp/pyproject.toml b/lzwcai_workflow_to_mcp/pyproject.toml
new file mode 100644
index 0000000..3dd7ec5
--- /dev/null
+++ b/lzwcai_workflow_to_mcp/pyproject.toml
@@ -0,0 +1,35 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "lzwcai-workflow-to-mcp"
+version = "0.1.8"
+description = "MCP server for executing business SQL queries with dynamic tool generation"
+readme = "README.md"
+requires-python = ">=3.13"
+license = {text = "MIT"}
+authors = [
+    {name = "lzwcai", email = "your-email@example.com"},
+]
+keywords = ["mcp", "sql", "executor", "server"]
+classifiers = [
+    "Development Status :: 3 - Alpha",
+    "Intended Audience :: Developers",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.13",
+]
+dependencies = [
+    "httpx>=0.28.1",
+    "mcp[cli]>=1.10.1",
+    "pypinyin>=0.53.0",
+]
+
+[project.scripts]
+lzwcai-workflow-to-mcp = "lzwcai_workflow_to_mcp.main:main"
+
+[tool.hatch.build.targets.wheel]
+packages = ["lzwcai_workflow_to_mcp"]
+
+[tool.hatch.build.targets.wheel.force-include]
+"lzwcai_workflow_to_mcp/businessQueries.json" = "lzwcai_workflow_to_mcp/businessQueries.json"