Smart Context MCP

Smart Context MCP

Reduces token consumption for AI coding agents by 50-70% through intelligent code context filtering, Git delta tracking, and local SQLite/Tree-sitter indexing.

Category
访问服务器

README

🦉 Smart Context MCP Server

Smart Context MCP is a local Model Context Protocol (MCP) server acting as a Universal Context Engine for AI Coding Agents (Cline, Roo Code, Claude Code, Cursor, Continue.dev, Codex, etc.).

It reduces token consumption for powerful LLMs (Strong LLMs) by 50%–70% by intelligently filtering code context, tracking Git Delta changes, and building high-performance local indexes using SQLite and Tree-sitter AST parsing.

Built following the ai-read-first specification & architecture principles for AI-assisted software development.


🌟 Key Features

  • Smart Token Optimization: Filters out irrelevant code files and sends only the precise context required for each prompt.
  • Git Delta Detection: Inspects git status and git diff to prioritize active workspace files.
  • SQLite Local Storage: Fast symbol and Repo Map querying (< 2ms) without high RAM overhead.
  • Multi-Project Management: Dynamically filters and tracks AST symbols per project.
  • Flexible Low AI Integration: Supports intent analysis using cheap or local AI providers (Ollama Nemotron, Groq Cloud, Anthropic Claude, OpenAI, Google Gemini Flash).
  • Web Dashboard UI: Built-in web administration dashboard at http://localhost:3333 for settings, tooltips, and SQLite cache exploration.

💾 Local Storage & Database Locations

Smart Context MCP persists user settings and indexed workspace data locally in your home directory across operating systems:

Data Type macOS / Linux Path Windows Path
⚙️ User Configuration ~/.smart-context/config.json %USERPROFILE%\.smart-context\config.json
📦 SQLite Cache Database ~/.smart-context/cache.db %USERPROFILE%\.smart-context\cache.db

(You can customize the cache database path anytime via the Web Dashboard settings modal).


⚡ Quick Start (Pre-built Release — No Build Required)

End-users do not need to install Node.js build tools or run npm:

  1. Download the pre-built release package: smart-context-mcp-v1.0.0.zip (from the release/ folder or GitHub Releases).
  2. Unzip smart-context-mcp-v1.0.0.zip into your desired folder.
  3. Configure your AI Client to point to the executable launcher smart-context-mcp:
{
  "mcpServers": {
    "smart-context": {
      "command": "/path/to/unzipped/smart-context-mcp-release/smart-context-mcp",
      "args": []
    }
  }
}

📂 Popular AI Client & VS Code Extension Config Locations

Here is where to find the MCP configuration file across Windows, macOS, and Linux for popular AI Coding Clients & VS Code extensions:

AI Client / Extension Operating System Standard Configuration File Path
Claude Desktop App macOS ~/Library/Application Support/Claude/claude_desktop_config.json
Windows %APPDATA%\Claude\claude_desktop_config.json
Linux ~/.config/Claude/claude_desktop_config.json
Cursor IDE macOS / Linux ~/.cursor/mcp.json (or Cursor Settings ➔ Features ➔ MCP)
Windows %USERPROFILE%\.cursor\mcp.json
Continue.dev (VS Code & JetBrains) macOS / Linux ~/.continue/config.json (under experimental.mcpServers)
Windows %USERPROFILE%\.continue\config.json
Cline Extension (VS Code) macOS ~/Library/Application Support/Code/User/globalStorage/saoudrizwan.claude-dev/settings/cline_mcp_settings.json
Windows %APPDATA%\Code\User\globalStorage\saoudrizwan.claude-dev\settings\cline_mcp_settings.json
Linux ~/.config/Code/User/globalStorage/saoudrizwan.claude-dev\settings\cline_mcp_settings.json
Roo Code Extension (VS Code) macOS ~/Library/Application Support/Code/User/globalStorage/rooveterinaryinc.roo-cline/settings/cline_mcp_settings.json
Windows %APPDATA%\Code\User\globalStorage\rooveterinaryinc.roo-cline\settings\cline_mcp_settings.json
Linux ~/.config/Code/User/globalStorage/rooveterinaryinc.roo-cline\settings\cline_mcp_settings.json
Claude Code CLI macOS / Linux / Windows ~/.claude.json (under mcpServers)
Codex / VS Code MCP macOS / Linux / Windows .vscode/mcp.json (Workspace settings)
Antigravity IDE macOS / Linux ~/.gemini/antigravity-ide/mcp_config.json
Windows %USERPROFILE%\.gemini\antigravity-ide\mcp_config.json

🖥️ Web Dashboard UI

When the MCP Server starts, the administration Web Dashboard is available at:

👉 http://localhost:3333

Features available on the Web Dashboard:

  • Project Filter: Select and view cached data for specific projects or across all repositories.
  • SQLite Cache Explorer: View cached files and parsed AST symbols with detailed hover tooltips (?).
  • Header Actions (⚙️ Settings & ❓ Guide Modals): Dynamically update Low AI Providers and view user guides in modal popups.

📖 Architecture & Standards

This project adheres to the ai-read-first standard for AI agent workflows.


🛠️ Developer Setup (For Modifying Source Code)

If you want to contribute or customize the source code:

# 1. Clone repository
git clone https://github.com/nghiale-soft/smart-context-mcp.git
cd smart-context-mcp

# 2. Install dependencies & build
npm install
npm run build

# 3. Create pre-built standalone ZIP release
npm run release

📜 Helper Scripts (scripts/)

  • ./scripts/release.sh: Builds TypeScript source and packages clean ZIP release (release/smart-context-mcp-v1.0.0.zip).

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选