Kortx
Enables AI assistants to consult advanced AI models for code analysis, strategic planning, and real-time web search with intelligent context gathering.
Kortx is a lightweight, open-source Model Context Protocol (MCP) server designed to empower AI assistants with specialized consultation capabilities. It integrates leading AI models like GPT-5 variants for strategic planning, code improvement, and problem-solving, alongside Perplexity Sonar for real-time web search with citations. A key feature is its intelligent, automated context gathering from codebases, making it invaluable for AI researchers, tool builders, and platform engineers looking to enhance AI assistant functionality and streamline development workflows.
Key Features
01Offers five specialized consultation tools: strategic planning, alternative solutions, copy improvement, problem solving, and web search.
021 GitHub stars
03Provides access to multiple GPT-5 variants (gpt-5, gpt-5-mini, gpt-5-nano, gpt-5-pro, gpt-5-codex) for diverse consultation tasks.
04Automatically gathers relevant code context using Serena, graph-memory, and cclsp MCPs.
05Designed for production with Docker support, non-root execution, sensitive-data redaction, and comprehensive structured logging.
06Integrates Perplexity Sonar models for real-time web search with comprehensive citations.
Use Cases
01Automating context gathering from codebases to improve AI-driven problem-solving and code improvement.
02Enhancing AI assistants (e.g., Claude Code, Copliot) with advanced code consultation and strategic planning capabilities.
03Integrating real-time web search and citation features into AI development and research workflows.