# Code Context Engine MCP server

> Indexes your codebase to provide AI clients with efficient, searchable code context while reducing token usage.

- Repository: https://github.com/elara-labs/code-context-engine
- Page: https://hostedmcp.org/servers/elara-labs/code-context-engine
- Transport: stdio
- Stars: 459
- License: MIT
- Latest version: 0.4.13

## Install

### Claude Code

```sh
claude mcp add code-context-engine -- uvx code-context-engine
```

### Cursor (`~/.cursor/mcp.json`)

```json
{
  "mcpServers": {
    "code-context-engine": {
      "command": "uvx",
      "args": [
        "code-context-engine"
      ]
    }
  }
}
```

### VS Code (`.vscode/mcp.json`)

```json
{
  "servers": {
    "code-context-engine": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "code-context-engine"
      ]
    }
  }
}
```

### Claude Desktop (`claude_desktop_config.json`)

```json
{
  "mcpServers": {
    "code-context-engine": {
      "command": "uvx",
      "args": [
        "code-context-engine"
      ]
    }
  }
}
```

### Codex (`~/.codex/config.toml`)

```toml
[mcp_servers.code-context-engine]
command = "uvx"
args = ["code-context-engine"]
```

### Gemini CLI (`~/.gemini/settings.json`)

```json
{
  "mcpServers": {
    "code-context-engine": {
      "command": "uvx",
      "args": [
        "code-context-engine"
      ]
    }
  }
}
```

## FAQ

### How much can I save on tokens?

The engine can provide up to 94% token savings by using indexed searches instead of re-reading entire files.

### Which editors are supported?

It supports a wide range of tools including Claude Code, VS Code, Cursor, Gemini CLI, and more.

### Does it work with Ollama?

Yes, if you have Ollama installed, CCE will auto-detect it at localhost:11434 and use nomic-embed-text.