Brasil MCP Server
Access 70 public Brazilian government datasets via 533 dedicated AI tools.
Data last scanned today · Reviewed today
Overview
This server provides structured integration with Brazilian public data, spanning 15 thematic areas such as economy, legislation, judicial records, elections, public health, and education. It utilizes over 500 individual tools to query APIs and local DuckDB-cached datasets, allowing AI assistants to cross-reference complex information like deputy spending, legislative voting history, and electoral filings. Built for performance, the server includes features for batch execution and query planning. While 66 of the integrated APIs are public and require no authentication, users can optionally configure specific keys for services that provide deeper data access.
Our verdict
With a security score of 39/100, this project is considered high risk. The score reflects that the project is community-maintained rather than an official government tool and lacks external editorial review, meaning users should exercise caution. Given that the repository has not received a commit in 90 days, it is best suited for experimental or research use rather than mission-critical applications.
- Supports 533 distinct tools for granular data retrieval
- Zero direct dependencies reduces supply chain surface
- Provides offline-capable local caching via DuckDB
- High risk score due to lack of official status
- Project development appears stalled with no recent commits
- Authentication methods for API access remain unreviewed
Tools
| Tool | Description | Risk |
|---|---|---|
| listar_features | Lista todas as features disponíveis no servidor. | low |
| recomendar_tools | Recomenda tools baseadas no contexto de consulta fornecido. | low |
| planejar_consulta | Cria planos de execução combinando múltiplas APIs. | low |
| executar_lote | Dispara múltiplas consultas em paralelo numa única chamada. | low |
| listar_datasets_disponiveis | Lista os datasets locais disponíveis para consulta via SQL. | low |
| ibge_listar_estados | Lista os 27 estados com codigo, nome, sigla e regiao | low |
| ibge_buscar_municipios | Municipios por codigo do estado | low |
| ibge_listar_regioes | Lista as 5 regioes do Brasil | low |
| ibge_consultar_nome | Frequencia de nomes pelo censo do IBGE | low |
| ibge_ranking_nomes | Nomes mais comuns por estado/municipio | low |
Compatibility
| Client | Local | Docker | Remote | Read-only |
|---|---|---|---|---|
| Claude Desktop | ||||
| Cursor | ||||
| VS Code | ||||
| Windsurf | ||||
| ChatGPT |
Frequently asked questions
›Do I need an API key to use mcp-brasil?
Most features do not require a key. Out of 70 total features, 66 work without keys, while 4 require free keys obtained through registration.
›Is this an official government service?
No, this is an independent project. It is not an official service of the Brazilian government or any of the institutions whose data it provides.
›How can I configure the server for Claude Desktop or VS Code?
You can add the server by specifying the command 'uvx' and the arguments '--from mcp-brasil python -m mcp_brasil.server' in your configuration file, along with optional environment variables for specific APIs.
›What is the function of the cross-referencing tool?
The 'planejar_consulta' tool allows the agent to create execution plans that combine multiple APIs to answer complex, multi-source queries.
›Are there usage restrictions I should be aware of?
Yes, all users must adhere to the provided Acceptable Use Policy (ACCEPTABLE_USE.md), and you should review the specific license requirements for each data source listed in SOURCES.md before commercial or journalistic use.
›Does the server support large datasets?
Yes, it supports large datasets via embedded DuckDB and local caching, which can be enabled as an opt-in configuration using environment variables.
Changelog
Badge
Maintain this server? Add the live badge to your README.