lakebrasil-mcp

lakebrasil-mcp

Enables agents to query 153 live Brazilian government data tools across federal and state sources, including economy, legislature, judiciary, elections, health, education, and more, reading directly from original APIs.

Category
访问服务器

README

<p align="center"> <img src="https://raw.githubusercontent.com/lakebrasil/data-br/main/assets/logo/wordmark.png" alt="lakebrasil" width="480"/> </p>

lakebrasil-mcp

License: MIT PyPI

153 tools, one MCP server, live Brazilian government data — federal AND state. Every tool reads straight from the original source, one period or entity at a time (never a bulk historical dump) — not from a mirrored copy. Point an agent (Claude, Cursor, any MCP client) at this server and it can look up the Selic rate, a CNPJ, a deputy's expenses, a court case, an active wildfire hotspot, a município's 2024 election results, a state budget line item, or a postal code, live.

Sibling project to lakebrasil, the open-source engine that aggregates 35 pipelines of historical data into an Iceberg lakehouse. This package is the live counterpart — most tools need no lake at all, hitting the government's own API/file for just the slice asked for; a handful (município reference data, CEPs, flags) query a bundled local dataset instead of the network on every call.

Install

uvx lakebrasil-mcp
# or
pip install lakebrasil-mcp

Add to an MCP client config (e.g. Claude Desktop, claude_desktop_config.json):

{
  "mcpServers": {
    "lakebrasil": {
      "command": "uvx",
      "args": ["lakebrasil-mcp"]
    }
  }
}

That's it — no account, no API key, for almost every tool. One exception: transparencia_* tools need PORTAL_TRANSPARENCIA_API_KEY (free, register at https://api.portaldatransparencia.gov.br/swagger-ui.html). (b3_* tools work unauthenticated too, but only for PETR4/VALE3/MGLU3/ITUB4 unless you also set BRAPI_TOKEN — brapi.dev's free-tier limit.)

What's in it (153 tools, 57 modules)

Domain Sources
Economy BACEN/SGS (Selic, IPCA, câmbio), BCB Olinda/PTAX, BNDES, Ipeadata, Comex Stat, CEMPRE/RAIS
Capital markets B3 quotes/dividends (via brapi.dev), CVM company filings/fatos relevantes/fund reports
Legislature Câmara dos Deputados, Senado Federal
Transparency & fiscal Portal da Transparência, TCU, SICONFI/Tesouro Nacional, Orçamento (execução), Emendas Parlamentares, FNDE/SIOPE
State governments Open-data (CKAN) portals — São Paulo, Rio Grande do Sul, Goiás, more as confirmed
Judiciary DataJud/CNJ, TST jurisprudência
Elections TSE (candidatos, resultados municipais), Meta Ad Library (anúncios políticos)
Geography & stats IBGE (localidades, agregados/SIDRA, regiões metropolitanas), município reference (57 fields incl. TSE/INEP/Tesouro), CEP lookup/search, flags
Procurement PNCP (compras públicas), TransfereGov
Environment INPE (queimadas), ANA (hidrologia)
Energy & transport ANEEL, ANTT, ANP (fuel prices), frota (DENATRAN/SENATRAN)
Aviation OpenSky
Public safety Atlas da Violência, Fórum Brasileiro de Segurança Pública
Health CNES, SISVAN/opendatasus, PNI/imunização, ANVISA, Banco de Preços em Saúde, Farmácia Popular, ANS (health plan beneficiaries)
Social programs BPC, Bolsa Família, Pé-de-Meia — municipal aggregates only (LGPD; never per-beneficiary data)
Education INEP Censo Escolar, MEC Censo da Educação Superior
Sanitation SINISA (current), SNIS (historical)
Municipal gazettes Querido Diário
News Agência Brasil
Utilities BrasilAPI (CEP, CNPJ, bancos, feriados, DDD)

Every tool's description documents its own caveats (rate limits, LGPD/ CC-BY-ND notes where relevant, which fields are unreliable) — read the tool description your MCP client surfaces before assuming a field means what it looks like it means.

Why public

The plan here is straightforward: build the widest possible community of people actually using this against real government data, learn from that usage, and let that shape whatever comes after — rather than guessing at a product behind a paywall from day one. If you hit a broken source, a wrong field, or a rate limit that shouldn't be there, open an issue.

Development

Ported from an original Rust implementation (rmcp + reqwest) to Python (mcp SDK's FastMCP + httpx). Every module follows the same shape: <source>/client.py (pure fetch/parse logic) and <source>/tools.py (@mcp.tool()-decorated functions).

pip install -e ".[dev]"
ruff check src/

License

MIT — see LICENSE.

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选