Brazilian Law Research MCP Server is an MCP server designed to enable agent-driven research on Brazilian law by scraping official sources. It provides scraping capacities to empower models to retrieve relevant legal precedents and statutes from Brazilian courts while emphasizing responsible use to avoid overloading official sources. The project requires git, uv (recommended) or Python 3.12+, and Google Chrome. To use, clone the repository, install dependencies via uv run patchright install, and configure an MCP client (example JSON) to point to the brlaw_mcp_server, invoking uv with directory path and serve command. Available tools include StjLegalPrecedentsRequest, TstLegalPrecedentsRequest, and StfLegalPrecedentsRequest to research precedents from STJ, TST, and STF that meet criteria. Development uses Ruff, BasedPyright, and Pytest. Language guidelines instruct Portuguese for content related resources/prompts to support non-dev users like lawyers and law students, while development remains in English. The project is MIT licensed.