Skip to main content
Glama

MCP DadosBR

CHANGELOG.md8.62 kB
# Changelog All notable changes to MCP DadosBR will be documented in this file. ## [Unreleased] ### Planned - **📱 Flutter/Dart Library**: Comprehensive Flutter library for mobile, web, and desktop applications - Cross-platform CNPJ and CEP lookup with the same reliable data sources - Smart caching with 60-second TTL and LRU eviction optimized for mobile - Circuit breaker pattern with exponential backoff for mobile networks - AI integration with LangChain Dart and LangGraph support - MCP server mode with HTTP/SSE transport for AI agent connectivity - Batch processing with Stream-based APIs for multiple document lookups - Comprehensive utility functions for validation, formatting, and offline verification - Platform-specific HTTP clients (dart:io for mobile/desktop, fetch for web) - Type-safe Dart classes mirroring the TypeScript interfaces - See [Flutter Library Documentation](docs/FLUTTER_LIBRARY.md) for details ## [0.2.0] - 2025-09-30 ### Added - **🔍 Web Search Tool (`cnpj_search`)**: Intelligent web search using DuckDuckGo - Support for advanced search operators (site:, intext:, intitle:, filetype:, etc.) - Rate limiting (1 second between requests) - Automatic caching of search results - Find lawsuits, documents, news, and online presence of companies - **🧠 Sequential Thinking Tool (`sequentialthinking`)**: Structured and iterative reasoning - Dynamic problem-solving with adaptable thinking process - Support for revisions and branches in reasoning - Colorized output for better visualization - Context tracking across multiple thought steps - Useful for complex analysis and investigations ### Enhanced - **Tool System**: Extended to support 4 tools (cnpj_lookup, cep_lookup, cnpj_search, sequentialthinking) - **Documentation**: Updated README with new tools and examples - **Architecture**: KISS approach - LLM orchestrates independent tools ### Technical - Added `duckduckgo-search` dependency for web search - Added `chalk` dependency for colorized thinking output - Zero cost for new features (DuckDuckGo is free) - Minimal changes to existing codebase ## [1.1.1] - 2025-09-28 ### Fixed - **MCP Schema Validation**: Fixed tool inputSchema format to use proper JSON Schema instead of Zod shapes - **Tool Registration**: Corrected `cnpj_lookup` and `cep_lookup` tools to have valid `type: "object"` schemas - **MCP Client Compatibility**: Resolved schema validation errors that prevented connection with MCP clients ### Updated - **Documentation**: Updated README.md with comprehensive integration guide for all AI coding tools - **Integration Examples**: Added configuration examples for Claude Desktop, Cursor, Windsurf, Continue.dev, VS Code - **Quick Setup Guide**: Created copy-paste configurations for immediate setup - **Version Sync**: Updated server version to match package.json version ## [1.1.0] - 2025-09-28 ### Added - **Cloudflare Workers Support**: Full deployment support for Cloudflare Workers with enhanced features - **Server-Sent Events (SSE)**: Real-time streaming MCP support via `/sse` endpoint - **ChatGPT Integration**: Direct integration with ChatGPT via OpenAPI schema and REST endpoints - **OAuth Endpoints**: Complete OAuth 2.0 flow for MCP connectors (`/.well-known/*`, `/oauth/*`) - **REST API Endpoints**: Direct HTTP access via `/cnpj/{cnpj}` and `/cep/{cep}` for ChatGPT actions - **OpenAPI Schema**: Auto-generated schema at `/openapi.json` for ChatGPT integration - **Simplified Integration**: Direct npm package usage without bridge files - **KV Storage Support**: Cloudflare KV integration for persistent caching across requests - **Enhanced Worker Features**: Health checks, CORS, DNS rebinding protection, and multi-environment support - **TypeScript Worker Config**: Separate `tsconfig.worker.json` for Cloudflare Workers compilation - **Wrangler Configuration**: Complete `wrangler.toml` with staging/production environments and KV namespaces ### Enhanced - **Multi-Transport Architecture**: Now supports stdio, HTTP, SSE, and REST API transports - **Documentation**: Added comprehensive ChatGPT integration guide and updated Cloudflare deployment docs - **Project Structure**: Enhanced organization with workers directory and bridge components - **Configuration**: Extended environment variable support for Cloudflare Workers deployment ### Features - **Global Edge Deployment**: Deploy to 300+ Cloudflare locations worldwide - **Zero Cold Starts**: Instant response times with Cloudflare Workers - **Free Tier Support**: 100k requests/day on Cloudflare's free tier - **Real-time Streaming**: SSE support for MCP agents with persistent connections - **ChatGPT Actions**: Direct integration with ChatGPT custom GPTs - **Remote MCP**: Bridge local MCP clients to remote Cloudflare Workers deployment ## [1.0.0] - 2025-09-27 ### Added - **Core MCP Server**: Initial release with Model Context Protocol support - **CNPJ Lookup Tool**: Brazilian company data via OpenCNPJ API integration - **CEP Lookup Tool**: Brazilian postal code data via OpenCEP API integration - **Dual Transport Support**: stdio (default) and HTTP Streamable transports - **Input Validation**: Zod-based validation and normalization for CNPJ/CEP formats - **Advanced Caching**: In-memory cache with 60-second TTL and LRU eviction (256 entries) - **Request Deduplication**: Promise-based deduplication prevents concurrent identical API calls - **Circuit Breaker**: API resilience with 5-failure threshold and 30-second recovery window - **Metrics Collection**: Built-in performance monitoring with request/cache/error tracking - **Configurable API Endpoints**: Support for custom CNPJ and CEP API URLs via configuration - **Configuration File Support**: `.mcprc.json` file for persistent configuration settings - **Authentication Headers**: Support for custom authentication headers in API requests - **Configuration Hierarchy**: Environment variables override config file settings - **Enhanced Logging**: API base URLs included in log output for better debugging - **URL Validation**: Automatic validation and normalization of custom API URLs - **Configuration Management**: Environment variables and `.mcprc.json` file support - **Comprehensive Error Handling**: Structured error responses with HTTP status mapping - **Timeout Protection**: 8-second AbortController timeout for all API requests - **Structured Logging**: One-line format with timestamp, tool, input, status, timing, and transport - **Simplified Deployment**: Focus on npm package and Cloudflare Workers - **Integration Tests**: Comprehensive test suite with real MCP protocol validation - **Advanced Features**: Graceful shutdown, CORS support, DNS rebinding protection ### Features - **Single-file Architecture**: All implementation in server.ts (480 lines of TypeScript) - **Zero Configuration**: Works out of the box with sensible defaults - **Enterprise Features**: Comprehensive error handling, monitoring, and deployment support - **High Performance**: Promise-based request deduplication, LRU caching, and circuit breaker patterns - **Flexible Deployment**: Supports both CLI tools (stdio) and web services (HTTP Streamable) - **Advanced Resilience**: Circuit breaker, request deduplication, and timeout protection - **Built-in Monitoring**: Automatic metrics collection and structured logging ### API Endpoints - `cnpj_lookup`: Query Brazilian company data by CNPJ number (via https://api.opencnpj.org/) - `cep_lookup`: Query Brazilian postal code data by CEP (via https://opencep.com/v1/) ### Configuration Options - **Transport Selection**: stdio (default) or HTTP Streamable transport - **Cache Configuration**: Configurable size (default: 256) and TTL (default: 60s) - **API Timeout**: Configurable timeout for external API calls (default: 8s) - **Circuit Breaker**: Configurable failure threshold (default: 5) and recovery window (default: 30s) - **HTTP Settings**: Port, CORS, DNS protection, allowed hosts/origins - **Metrics & Logging**: Enable/disable metrics collection and configure log levels - **Configuration Sources**: Environment variables, `.mcprc.json` file, or built-in defaults ### Testing - Manual testing scripts for both transports - Integration test suite with real MCP protocol validation - Comprehensive testing suite - Performance benchmarking capabilities ### Documentation - Comprehensive README with usage examples - API documentation with request/response formats - Configuration guide with all available options - Deployment guide for different environments - Troubleshooting guide and FAQ

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/cristianoaredes/mcp-dadosbr'

If you have feedback or need assistance with the MCP directory API, please join our Discord server