openrouter-deep-research-mcp
If you are the rightful owner of openrouter-deep-research-mcp and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.
OpenRouter Agents MCP Server is a sophisticated orchestration system for AI-powered research, utilizing multiple specialized agents and models.
The OpenRouter Agents MCP Server is a Model Context Protocol (MCP) server designed to enhance AI-powered research capabilities. It implements a multi-agent orchestration system that leverages various OpenRouter models to conduct comprehensive research. The server supports both STDIO and HTTP/SSE transports, enabling seamless integration with different platforms. It features a vector embedding database for semantic knowledge storage and employs round-robin load balancing to optimize research task distribution. The server is equipped with an adaptive fallback system to ensure continuity in research even when primary models fail. Recent enhancements include cross-model resilience, dynamic caching, and improved user feedback mechanisms. The server is compatible with tools like Cline in VS Code and the Claude Desktop App, providing enterprise-grade research capabilities in a user-friendly package.
Features
- Multi-Agent Orchestration: Hierarchical system with planning, research, and context agent roles.
- Vector Embedding Database: PGLite with pgvector for semantic knowledge storage.
- Round-Robin Load Balancing: Distributes research tasks across different models for optimal results.
- Adaptive Fallback System: High to low-cost model degradation when primary research fails.
- Cross-Model Resilience: Comprehensive error handling keeps research flowing despite individual model failures.
Tools
Basic Tool Testing
Test whether all MCP tools work properly
MCP Server Testing
Test the MCP server implementation, including all transfer options
Research Agent Testing
Test core research proxy functions, including actual OpenRouter API calls
