Spec Driven Development
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
作者提供的说明(英文原文)
Quadruple consensus MCP server for code validation using Codex, Gemini, and Qwen
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
打开 Claude Desktop → Settings → Developer → Edit Config,把下面的内容粘贴到 claude_desktop_config.json
{
"mcpServers": {
"tetrad": {
"command": "npx",
"args": [
"-y",
"@samoradc/tetrad"
]
}
}
}在终端中运行这条命令
claude mcp add tetrad -- npx -y @samoradc/tetrad粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"tetrad": {
"command": "npx",
"args": [
"-y",
"@samoradc/tetrad"
]
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"tetrad": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@samoradc/tetrad"
]
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个服务器在你的电脑上运行,而 ChatGPT 只能连接互联网上的服务器。请选择其他应用。
你的选择通过 Cookie 保存在这台设备上,无需账户。
这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月7日
注册名称已通过 GitHub 账户 SamoraDC 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
它在你的电脑上运行;我们只能读取互联网端点的工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
GitHub 仓库最后更新于–,最新版本(0.1.12)于9个月前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 Node.js。
同一分类中有:Spec Driven Development, Remote Desktop Commander, Master CSS
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.
Inspect Master CSS classes, generated CSS, and project stylesheets from your AI client.
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
Quadruple Consensus MCP Server for Claude Code
Tetrad is a high-performance MCP (Model Context Protocol) server written in Rust that orchestrates three AI-powered CLI code evaluation tools (Codex, Gemini CLI, Qwen) to validate all code produced by Claude Code.
The system implements a quadruple consensus protocol where no code or plan is accepted without approval from four intelligences: the three external evaluators + Claude Code itself.
init, serve, status, doctor, config, etc.)# Via npm (recommended - works seamlessly with Claude Code)
npm install -g @samoradc/tetrad
cargo install tetrad sudo cp ~/.cargo/bin/tetrad /usr/local/bin/
npx @samoradc/tetrad init
This will:
tetrad.toml configuration file.tetrad/ directory for the database.tetrad/ to your .gitignoreTetrad requires at least one of the following AI CLI tools. Good news: you don't need separate API keys if you have existing subscriptions!
npm install -g @openai/codex
Authentication options:
| Method | Requirements | API Key Needed? |
|---|---|---|
| ChatGPT Login | ChatGPT Plus, Pro, Business, Edu, or Enterprise subscription | No |
| API Key | OpenAI API account with credits | Yes |
# Option 1: Login with your ChatGPT subscription (recommended)
codex login
export OPENAI_API_KEY="your-openai-key"
Note: For headless/remote servers, use
codex login --device-authand follow the device code flow.
📖 Codex CLI Authentication Docs
npm install -g @google/gemini-cli
Authentication options:
| Method | Requirements | API Key Needed? | Free Tier |
|---|---|---|---|
| Google Login | Personal Google account | No | 60 req/min, 1,000 req/day |
| API Key | Google AI Studio account | Yes | Same limits |
# Option 1: Login with Google account (recommended)
gemini auth login
export GEMINI_API_KEY="your-gemini-key"
Free Tier: With a personal Google account, you get access to Gemini 2.5 Pro with 1M token context window at no cost!
📖 Gemini CLI Authentication Docs
npm install -g qwen-code
Authentication options:
| Method | Requirements | API Key Needed? | Free Tier |
|---|---|---|---|
| DashScope API | Alibaba Cloud account | Yes | 2,000 req/day |
| Coding Plan | Qwen Coding subscription | Yes (different format) | Included |
# Get your API key from DashScope console
export DASHSCOPE_API_KEY="your-dashscope-key"
export OPENAI_BASE_URL="https://dashscope-intl.aliyuncs.com/compatible-mode/v1"
Regional Endpoints:
- Singapore:
dashscope-intl.aliyuncs.com
- Virginia:
dashscope-us.aliyuncs.com
- Beijing:
dashscope.aliyuncs.com
# Check Tetrad version
tetrad version
tetrad status
tetrad doctor
# Add Tetrad as MCP server (all projects)
claude mcp add --scope user tetrad -- npx @samoradc/tetrad serve
claude mcp list
Alternative installation methods
# Current project only
claude mcp add tetrad -- npx @samoradc/tetrad serve
claude mcp add --scope user tetrad -- tetrad serve
Note: Tetrad is published on the MCP Registry for discoverability and documentation.
Create or edit .mcp.json in your project root:
{
"mcpServers": { "tetrad": { "type": "stdio", "command": "npx", "args": ["@samoradc/tetrad", "serve"], "env": { "OPENAI_API_KEY": "${OPENAI_API_KEY}", "GOOGLE_API_KEY": "${GOOGLE_API_KEY}", "DASHSCOPE_API_KEY": "${DASHSCOPE_API_KEY}" } } } }
Or for global user configuration in ~/.claude.json (mcpServers section):
{
"mcpServers": { "tetrad": { "type": "stdio", "command": "npx", "args": ["@samoradc/tetrad", "serve"] } } }
When you ask Claude Code to write code, Tetrad automatically validates it:
You: "Create a function in Rust that calculates the average of a vector"
Claude Code:
Tetrad Response: ┌─────────────────────────────────────────────┐ │ DECISION: PASS ✓ │ │ Score: 92/100 │ │ Consensus: Yes (3/3 approved) │ │ │ │ Votes: │ │ • Codex: Pass (95) │ │ • Gemini: Pass (90) │ │ • Qwen: Pass (92) │ │ │ │ Suggestions: │ │ - Consider handling empty vector case │ └─────────────────────────────────────────────┘
Claude Code: Saves the approved code
Tetrad Response:
┌─────────────────────────────────────────────┐ │ DECISION: BLOCK ✗ │ │ Score: 26/100 │ │ Consensus: Yes (3/3 rejected) │ │ │ │ Votes: │ │ • Codex: Fail (30) - division by zero │ │ • Gemini: Fail (25) - no error handling │ │ • Qwen: Fail (25) - unsafe operation │ │ │ │ Issues: │ │ - Division by zero not handled │ │ - Missing input validation │ │ - No Result/Option return type │ └─────────────────────────────────────────────┘
Claude Code: Fixes the code and resubmits
tetrad - Quadruple Consensus CLI for Claude Code
COMMANDS: init Initialize configuration in current directory serve Start the MCP server (used by Claude Code) status Show CLI status (codex, gemini, qwen) config Configure options interactively doctor Diagnose configuration issues version Show version evaluate Evaluate code manually (without MCP) history Show evaluation history from ReasoningBank export Export patterns from ReasoningBank import Import patterns into ReasoningBank
OPTIONS: -c, --config Configuration file (default: tetrad.toml) -v, --verbose Verbose mode -q, --quiet Quiet mode -h, --help Show help
When running as MCP server, Tetrad exposes 6 tools:
| Tool | Description |
|---|---|
tetrad_review_plan | Review implementation plans before coding |
tetrad_review_code | Review code before saving |
tetrad_review_tests | Review tests before finalizing |
tetrad_confirm | Confirm agreement with received feedback |
tetrad_final_check | Final verification before commit |
tetrad_status | Check health of evaluators |
1. Claude Code generates plan → tetrad_review_plan → Feedback
Claude Code → MCP Protocol (stdio) → Tetrad Server (Rust)
│ ┌─────────────────────┼─────────────────────┐ ▼ ▼ ▼ Codex CLI Gemini CLI Qwen CLI (syntax) (architecture) (logic) │ │ │ └─────────────────────┼─────────────────────┘ ▼ Consensus Engine │ ┌─────────────┴─────────────┐ ▼ ▼ LRU Cache ReasoningBank (results) (SQLite) RETRIEVE→JUDGE→DISTILL→CONSOLIDATE
| Executor | CLI | Specialization |
|---|---|---|
| Codex | codex exec --json | Syntax and code conventions |
| Gemini | gemini -o json | Architecture and design |
| Qwen | qwen | Logic bugs and correctness |
| Rule | Requirement | Use Case |
|---|---|---|
| Golden | Unanimity (3/3) | Critical code, security |
| Strong | 3/3 or 2/3 with high confidence | Default |
| Weak | Simple majority (2/3) | Rapid prototyping |
The ReasoningBank is a continuous learning system that stores and consolidates code patterns:
RETRIEVE → JUDGE → DISTILL → CONSOLIDATE
│ │ │ │ │ │ │ └─ Merge similar patterns │ │ └─ Extract new patterns │ └─ Evaluate code with context └─ Search for relevant patterns
只显示了开头部分;其余内容请在仓库中阅读