Awesome MCP › Software Development

Context-Engine-AI/Context-Engine

⭐ 400 Python added to this list on 2026-01-31 repository created 2025-10-08

Context-Engine MCP is an open-core, self-improving code search platform designed to enhance code exploration and understanding through agentic context compression. It leverages the Model Context Protocol (MCP) to provide a unified interface for indexing, searching, and memory services, making it highly suitable for developers and AI coding assistants. The project offers a command-line interface (CLI) called `ctx` that simplifies setup, service management, and querying, enabling users to quickly index their codebase and perform semantic searches or get detailed answers with citations. The architecture of Context-Engine is modular and layered, consisting of a client layer (IDE and VS Code extension), an MCP layer (search and index services), a processing layer (embedding service, LLM decoder, learning worker, and file watcher), and a storage layer (vector database Qdrant, Redis cache, and filesystem). This design supports both local mode for personal use and enterprise-ready features such as built-in authentication, session management, and workspace-aware queries. Context-Engine integrates with popular AI coding assistants like Codex, Claude, Gemini, and Augment by providing specialized agent skills that teach these assistants to utilize MCP tools effectively, replacing traditional code exploration methods like grep or find. The system supports multiple communication protocols including SSE and HTTP for streaming and remote procedure calls. The project also includes comprehensive documentation covering setup, configuration, API references, and troubleshooting. It supports advanced features like automatic collection injection, adaptive reranking, and delta bundle uploads. Benchmarks demonstrate strong performance in dense retrieval tasks on code search datasets, outperforming traditional methods. Overall, Context-Engine MCP is a sophisticated, scalable, and extensible platform that leverages MCP to improve code search and AI-assisted coding workflows, making it a valuable tool for developers and AI integration.

https://github.com/Context-Engine-AI/Context-Engine

adaptive-rerankingagentic-context-compressionaiai-agentsai-coding-assistantsai-integrationaugmentauthenticationclaudeclicode-explorationcode-searchcodexcompressioncontextcontext-enginectxcursordecoderdense-retrievalembedding-serviceenterprise-featuresgeminiglmhttpllmllm-decoderllm-inferencemcpmcp-servermodel-context-protocolollama-apiopen-coreqdrantredisrefragself-improvingsemantic-searchsession-managementssestreaming-protocolsvector-databaseworkspace-aware-queries