Dados do Brasil
Dados do Brasil is an independent open-source project that connects AI agents to Brazilian public data. Its MCP server provides structured, natural-language access to official sources covering economics, legislation, transparency, justice, elections, health, education, environment, public safety and more.
At a Glance
- Developers building MCP-compatible AI agents
- Researchers and data scientists
- Journalists and fact-checking teams
- OSINT practitioners
- +3 more
AI Tools by Dados do Brasil
(1)
mcp-brasil
MCP Server for Brazilian Data
Discussions
No discussions yet
Be the first to start a discussion about Dados do Brasil
Latest News
Weekly source-health canary and startup/deployment hygiene improvements merged
v0.14.0 released with DuckDB dataset infrastructure and multi-year TSE datasets
v0.12.0 added configurable authentication, expanded dados.gov.br, education and epidemiology features, and new deployment/connector guides
v0.11.0 introduced configurable authentication strategy
Products & Services
Python open-source Model Context Protocol server, installable with pip or uv, exposing tools, resources and prompts for Brazilian public data. The current README describes 533 tools, 131 resources, 102 prompts, 70 features and 15 thematic areas, with 70 public data sources.
The installable package distributed as mcp-brasil; it supports local stdio use and HTTP transport through FastMCP, and can be configured for Claude Desktop, Claude Code, VS Code/Cursor and Google Antigravity.
Opt-in local cache/query layer for large datasets, including SPU SIAPA, TSE 2014-2024 candidates/assets/voting/social-network/FEFC data, ANP fuel prices, INEP education data, ISP-RJ safety data and ANAC aviation data.
Quick-start guides, source/acceptable-use documentation, feature catalog and use-case guides explaining installation, client configuration, API keys and querying patterns.
Market Position
Dados do Brasil is positioned as a broad, open-source aggregation layer for Brazilian public data rather than a single-domain data product: it standardizes dozens of official APIs behind one MCP server and supports cross-source queries. This differentiates it from narrower MCP servers focused on a handful of datasets or one domain, while its MIT license and self-hosting model distinguish it from proprietary hosted data platforms.
Leadership
Founders
Jonatas Soares
Creator and lead maintainer associated with the GitHub handle jxnxts; his profile describes him as a Product & Tech Lead, technical consultant, and specialist in AI agents and HPC. Independent coverage identifies him as the creator of mcp-brasil.
Executive Team
Jonatas Soares
Creator and lead maintainer
GitHub profile: Product & Tech Lead, technical consultant, AI agents and HPC. He is the primary maintainer/committer associated with the project and is identified by independent coverage as its creator.
Founding Story
The project was started as an independent community open-source effort to make Brazil's fragmented public-data APIs usable by AI agents through the Model Context Protocol, rather than requiring users to learn and integrate each endpoint separately. Its initial public-launch changelog describes a rewrite of the README for public launch and a single server aggregating many government data sources.
Business Model
Revenue Model
Open-source MIT-licensed software distributed for free; users install and run it locally or deploy it themselves. The server uses third-party public APIs, with some optional free API-key registrations for higher-rate or protected sources.
Target Markets
- Developers building MCP-compatible AI agents
- Researchers and data scientists
- Journalists and fact-checking teams
- OSINT practitioners
- Civic-tech and transparency organizations
- Public-sector and policy analysts
- Government-data research and public-interest analysis
- OSINT investigations and fact checking
- Legislative research involving bills, votes, deputies and senators
- Economic and financial analysis using Selic, IPCA, exchange-rate and other series
- Transparency, procurement, contracts, spending and sanctions research
- Judicial and case-law research