Context-Engine-AI/Context-Engine

Context-Engine MCP - Agentic Context Compression Suite

What it solves

AI coding assistants often struggle to find the right code in large or multi-repository codebases. They rely on fragile grep or read-the-file approaches that waste tokens and miss cross-file meaning. Context Engine gives these assistants a reliable, semantic layer: indexed code that can be searched by behavior, concept, error text, or symbol relationships, with exact line ranges and verified source.

How it works

Context Engine indexes your codebase and exposes more than 30 MCP (Model Context Protocol) tools. The assistant learns to use them via skill files installed per editor (Claude Code, Cursor, Codex, Windsurf, Augment, Gemini, or any custom-instructions assistant). A CLI bridge (ctxce) connects your workspace, runs as a daemon, watches for changes, and serves MCP over stdio or HTTP. Key tools include repo_search for semantic queries, context_answer for cross-file how/why explanations with citations, symbol_graph for definitions/callers/callees/references, graph_query for multi-hop impact analysis, cross_repo_search for multi-repo tracing, memory_store/memory_find for persistent context, and pattern search for structural idioms.

Who it’s for

Developers using AI coding assistants (Claude Code, Cursor, Codex, Windsurf, etc.) who work in real-world codebases — especially large, multi-file, or multi-repository projects — and want the assistant to understand the code before editing, renaming, or answering questions about it.

Highlights

  • 30+ MCP tools covering semantic search, symbol graph navigation, cross-repo search, and persistent memory.
  • Skills that install into Claude Code, Cursor, Codex, Windsurf, Augment, Gemini, or any assistant via a single SKILL.md.
  • Batch queries (batch_search, batch_symbol_graph, batch_graph_query) that save 75%+ on tokens.
  • CLI bridge with daemon mode and file watching (ctxce connect), plus stdio and HTTP MCP serving.
  • Guides the assistant to use search first instead of grep, treating results as already-read verified source.

Related

  • Project
    forloopcodes/contextplus

    An MCP server that turns large codebases into searchable, hierarchical feature graphs using RAG, AST parsing, and spectral clustering for high-accuracy AI coding context.

  • Project
    CodeBendKit/codeseek

    A code intelligence CLI tool that combines AST-based call graph analysis and hybrid semantic search to help developers and AI agents navigate large codebases.

  • Project
    SammySnake-d/fast-context-mcp

    An MCP server that provides AI-driven semantic code search for AI clients, using Windsurf's Devstral API to iteratively find relevant code snippets via local commands.

  • Project
    hitmux/hitmux-context-engine

    A semantic context search engine for MCP clients that indexes local directories into vector storage to help AI agents find relevant code and documentation.

  • Project
    giancarloerra/SocratiCode

    A local, zero-config codebase intelligence engine that provides AI assistants with deep semantic understanding, dependency graphs, and symbol-level impact analysis for large repositories.