code-context-engine

Overview

Code Context Engine is a Python-based MCP server that creates a searchable index of your codebase, allowing AI coding assistants to retrieve relevant code snippets instead of reading entire files. It provides significant token savings (94% according to their benchmarks) while keeping all data local. The server offers 9 MCP tools for context search, session recall, and index management, working with multiple AI coding environments like Claude Code, VS Code, Cursor, and Gemini CLI.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• context_search

  Hybrid vector + BM25 search with graph expansion to find relevant code chunks

  Try:Find payment processing functions in the codebaseTry:Show me how authentication is implementedTry:Search for functions that handle user registration
• expand_chunk

  Retrieve full source code for a compressed result

  Try:Show me the full implementation of that functionTry:Expand the code chunk for payment processingTry:Get the complete source for the authentication module
• related_context

  Find code related via graph edges (calls, imports)

  Try:What files import the authentication module?Try:Show me functions called by payment processingTry:Find modules that use this database schema
• session_recall

  Recall decisions and context from past sessions

  Try:What decisions were made about authentication?Try:Show me previous work on payment processingTry:What were the reasons for choosing this architecture?
• record_decision

  Save a decision for future reference

  Try:Record decision: use JWT for authenticationTry:Note that we're using Stripe for payment processingTry:Save the decision to use PostgreSQL as the database
• record_code_area

  Record which files were worked on for future reference

  Try:Record that we're working on the payment moduleTry:Note that we modified user authentication codeTry:Remember that we're updating the database schema
• index_status

  Check the freshness and status of the code index

  Try:Is the code index up to date?Try:Show me the status of the indexTry:Check if the index needs rebuilding
• reindex

  Re-index a specific file or the entire project

  Try:Re-index the payment processing moduleTry:Update the index for all modified filesTry:Rebuild the entire project index
• set_output_compression

  Adjust response verbosity level

  Try:Set compression to standardTry:Use maximum compression for responsesTry:Switch to lite mode for responses

Comparable tools

Installation

# Install with uv
uv tool install code-context-engine

# Or with pipx
pipx install code-context-engine

# Initialize your project
cd /path/to/your/project
cce init                              # or: cce init --agent all

{
  "mcpServers": {
    "code-context-engine": {
      "command": "uv",
      "args": ["run", "code-context-engine", "mcp"]
    }
  }
}

FAQ

What embedding backends does CCE support?

CCE auto-detects the best available backend. It uses nomic-embed-text with Ollama, or includes fastembed + ONNX Runtime for local offline embedding when you install the [local] extra.

Does CCE work with multiple AI coding assistants simultaneously?

Yes. CCE can index your codebase once and work with Claude Code, VS Code, Cursor, Gemini CLI, Codex CLI, OpenCode, and Tabnine simultaneously with a single index.

Compare code-context-engine with