Stitch Skills

為什麼要用

核心特性

Agent Skills library for the Stitch MCP server - design-to-code conversion patterns for AI coding agents.
Maintained by google-labs-code (official)
Written in TypeScript
5665+ GitHub stars at last check
Open source on GitHub

即時演示

實際使用效果

stitch-skills.replay ▶ 就緒

0/0

安裝

選擇你的客戶端

~/Library/Application Support/Claude/claude_desktop_config.json · Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "stitch-skills": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/google-labs-code/stitch-skills",
        "~/.claude/skills/stitch-skills"
      ],
      "_inferred": true
    }
  }
}

開啟 Claude Desktop → Settings → Developer → Edit Config。儲存後重啟應用。

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "stitch-skills": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/google-labs-code/stitch-skills",
        "~/.claude/skills/stitch-skills"
      ],
      "_inferred": true
    }
  }
}

Cursor 使用與 Claude Desktop 相同的 mcpServers 格式。專案級設定優先於全域。

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "stitch-skills": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/google-labs-code/stitch-skills",
        "~/.claude/skills/stitch-skills"
      ],
      "_inferred": true
    }
  }
}

點擊 Cline 側欄中的 MCP Servers 圖示，然後選 "Edit Configuration"。

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "stitch-skills": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/google-labs-code/stitch-skills",
        "~/.claude/skills/stitch-skills"
      ],
      "_inferred": true
    }
  }
}

格式與 Claude Desktop 相同。重啟 Windsurf 生效。

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "stitch-skills",
      "command": "git",
      "args": [
        "clone",
        "https://github.com/google-labs-code/stitch-skills",
        "~/.claude/skills/stitch-skills"
      ]
    }
  ]
}

Continue 使用伺服器物件陣列，而非映射。

~/.config/zed/settings.json

{
  "context_servers": {
    "stitch-skills": {
      "command": {
        "path": "git",
        "args": [
          "clone",
          "https://github.com/google-labs-code/stitch-skills",
          "~/.claude/skills/stitch-skills"
        ]
      }
    }
  }
}

加入 context_servers。Zed 儲存後熱重載。

claude mcp add stitch-skills -- git clone https://github.com/google-labs-code/stitch-skills ~/.claude/skills/stitch-skills

一行命令搞定。用 claude mcp list 驗證，claude mcp remove 移除。

使用場景

實戰用法： Stitch Skills

Use Stitch Skills for its primary workflow

👤 Developers using AI coding agents ⏱ ~10 min beginner

何時使用： You need the capability: Agent Skills library for the Stitch MCP server - design-to-code conversion patterns for AI coding agents.

前置條件

Skill installed — git clone https://github.com/google-labs-code/stitch-skills ~/.claude/skills/stitch-skills

步驟

Install

Install Stitch Skills following the README at https://github.com/google-labs-code/stitch-skills✓ 已複製

→ Skill appears in your client's available tools/skills
Invoke

Use Stitch Skills to complete the task described in its docs.✓ 已複製

→ Agent calls the relevant tool/skill and returns a result

結果： Working integration with the capabilities advertised by the project.

注意事項

README is the source of truth; details here may lag upstream — Check https://github.com/google-labs-code/stitch-skills for the latest setup steps and tool list