mcp-law-server
Provides vector similarity search for law titles, clauses, and audit items, enabling semantic retrieval.
Provides relational query for hierarchical and categorical law data, supporting structured lookups and filtering.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@mcp-law-serverfind regulations related to government procurement audit"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
mcp-law-server
审计法规智能检索 MCP Server —— 面向审计 Agent 的法规/条款/审计事项检索与核实服务。
Milvus 向量检索(法规名 / 条款 / 审计事项)+ MySQL 关系查询(层级 / 分类)+ Agent 核实
技术栈
层 | 选型 |
语言 / 运行时 | Python 3.12 |
包管理 / 构建 | uv + hatchling(src-layout) |
MCP 框架 | FastMCP(HTTP transport,uvicorn/ASGI) |
MySQL 访问 | SQLAlchemy 2.x async + asyncmy(双库:MAIN / TOOLS) |
向量库 | Milvus(pymilvus),库 |
Embedding | BGE-large-zh-v1.5,1024 维(本地 FlagEmbedding 推理,按需安装) |
配置 | pydantic-settings + |
编排(平台侧) | Hermes YAML workflow + Agent prompts( |
Related MCP server: hivelaw
目录结构
mcp_law_server/
├── pyproject.toml # 元数据 / 依赖 / 入口 / 工具配置
├── .env.example # 环境变量模板(复制为 .env 填写)
├── Makefile # 常用命令封装
├── src/mcp_law_server/ # ── 应用代码(src-layout,可 pip install -e .)──
│ ├── main.py # 入口
│ ├── config/ # 配置层(无内部依赖)—— settings 单例
│ ├── models/ # 数据模型层(仅依赖 config)—— enums / schemas / errors
│ ├── db/ # MySQL 数据访问层 —— connection + 4 个 DAO
│ ├── vector/ # Milvus 向量访问层 —— client + 3 个 search
│ ├── embedding/ # Embedding 服务层 —— bge_embedder
│ ├── cache/ # 缓存层 —— registry_cache
│ ├── tools/ # MCP 工具层 —— 9 个 tool handler(核心业务)
│ └── server/ # MCP Server 注册与启动 —— mcp_app
├── hermes/ # ── Hermes 编排(平台侧,YAML + prompts,非 Python)──
├── scripts/ # ── 运维脚本(向量化入库 / 阈值校准 / 数据探查)──
└── tests/ # ── 测试(db / vector / tools / integration)──依赖方向(从下到上,禁止反向引用):
config → models → {db, vector, embedding, cache} → tools → server → main
快速开始
# 1. 安装核心 + 开发依赖(uv 会自动准备 Python 3.12 解释器)
uv sync
# 2. (按需)安装 ML 依赖(FlagEmbedding / torch,体积较大)
uv sync --extra ml
# 3. 配置环境变量
cp .env.example .env # 然后编辑 .env 填写真实值
# 4. 启动 MCP HTTP server
uv run mcp-law-server # 等价:make runWindows 若未安装
make,直接用上面的uv命令即可,Makefile仅做封装。
常用命令
命令 | 说明 |
| 安装核心 + 开发依赖 |
| 额外安装 ML 依赖 |
| 启动 MCP server |
| 运行测试 |
| 代码质量 |
| 清理缓存与 venv |
环境变量
见 .env.example,字段与 src/mcp_law_server/config/settings.py 一一对应。
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Latest Blog Posts
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/oranger07/mcp_law_server'
If you have feedback or need assistance with the MCP directory API, please join our Discord server