brasil-data-mcp is an open-source MCP server that exposes Brazilian public data to AI clients as tools, allowing users to obtain structured data by asking questions in natural language. It is powered by BrasilAPI, requires no key and no authentication, and the data comes from official sources.
Purpose: to connect AI clients with Brazilian public data, avoiding manual scraping or writing HTTP integrations. Users can directly ask about CNPJ company information, CEP addresses, bank codes, holidays, area codes, ISBN, economic indicators, exchange rates, CVM brokers, IBGE states/municipalities, .br domain availability, and more.
Main capabilities: provides 15 tools, covering CNPJ, CEP, banks, holidays, DDD, ISBN, economic indicators (SELIC/CDI/IPCA), exchange rates (PTAX/BACEN), CVM brokers, IBGE states and municipalities, and .br domain availability. It also provides 2 MCP prompts: analise-cnpj and panorama-economico, for guided workflows. Advanced data sources can optionally be enabled (via the CPFCNPJ_TOKEN environment variable); once enabled, consultar_cnpj can query official data in real time (D+0), and two new tools, consultar_cpf and consultar_inscricao_estadual, are added; if the advanced source fails, it automatically falls back to BrasilAPI.
Usage: run via npx -y brasil-data-mcp, with no global installation required. In Claude Desktop, edit the configuration file (macOS: ~/Library/Application Support/Claude/claude_desktop_config.json; Windows: %APPDATA%\Claude\claude_desktop_config.json), and restart to use it. In Cursor, create or edit .cursor/mcp.json in the project root. In Claude Code, you can add it using the command claude mcp add brasil-data -e CPFCNPJ_TOKEN=seu-token -- npx -y brasil-data-mcp. Advanced dataThe source token must be obtained at cpfcnpj.com.br/dev, and each query consumes account credits.
Use cases: suitable for Brazilian developers querying public data in daily LLM use, such as looking up company registration information, addresses, banks, holidays, economic indicators, exchange rates, domain availability, and more. It is also suitable for scenarios that require integrating Brazilian public data into AI workflows.
Limitations: by default, data comes from BrasilAPI and may not be real-time; advanced data sources require a paid token, and each query consumes credits. The project is open source under the MIT license, with no malicious rate limiting. FIPE data is not yet supported, pending upstream stabilization.
