lease-qa-mcp
MCP server exposing lease Q&A via the lease_qa tool, querying a LeaseClear API to return grounded answers about specific lease documents.
README
leaseclear-mcp
A standalone MCP server that exposes lease Q&A as one tool: lease_qa. It asks hosted LeaseClear a question about lease terms and returns a grounded answer.
This is a thin stdio process. By default it calls the public LeaseClear API. Override LEASECLEAR_API_URL only if you run your own instance.
Install
Requires Python 3.12+.
uvx leaseclear-mcp
Or:
pip install leaseclear-mcp
leaseclear-mcp
That starts the MCP server on stdio.
Configure
| Variable | Required | Meaning |
|---|---|---|
LEASECLEAR_API_URL |
no | LeaseClear API URL. Defaults to https://leaseclear-production.up.railway.app. Set this for a local or private instance. |
LEASECLEAR_API_KEY |
no | LeaseClear API key (POST /auth/api-key while logged in). Query-only: asks questions as that user. Omit to use the public demo corpus. |
If the key is set, this server sends it on /query. If omitted, it uses /auth/demo (no account needed).
From a git checkout you can copy .env.example to .env. After pip / uvx, set overrides in the MCP host env block instead.
Cursor / Claude Desktop example
{
"mcpServers": {
"lease-qa": {
"command": "uvx",
"args": ["leaseclear-mcp"]
}
}
}
To use your own account (or a self-hosted API), add env:
{
"mcpServers": {
"lease-qa": {
"command": "uvx",
"args": ["leaseclear-mcp"],
"env": {
"LEASECLEAR_API_KEY": "lc_…"
}
}
}
}
Tool: lease_qa
| Arguments | question (string) — one neutral question about lease terms |
| Metadata | document_ids (list of UUIDs, optional) — which leases to query; not an LLM-visible argument |
| Returns | { "answer": "..." } — LeaseClear’s grounded answer (or that the lease is silent) |
If the host passes document_ids, the question is scoped to those leases. If omitted, LeaseClear searches all leases on the authenticated account. The model only chooses the question; the host chooses whether to scope.
Errors
Failures surface as MCP tool errors, including:
- invalid
document_idsin metadata - empty question
- LeaseClear auth or query failures
- LeaseClear unreachable
Development
git clone https://github.com/amadeuserras/leaseclear-mcp.git
cd leaseclear-mcp
uv sync
cp .env.example .env
uv run leaseclear-mcp
uv run ruff check .
uv run pyright
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。