Skip to main content
Glama
Sign In
Waferlock Robot MCP

Waferlock Robot MCP

by jasonlcs
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Remote

Waferlock MCP CLI (Public)

這個專案提供一個獨立的 CLI，透過 Waferlock Robot MCP 在 Heroku 上的公開 REST API，將資料橋接到支援 Model Context Protocol (MCP) 的客戶端（例如 ChatGPT Desktop）。使用者只需要 API URL 與 Token，即可在本地啟動 stdio 版 MCP 伺服器，不需暴露 AWS 憑證。

安裝與使用

1. 取得程式碼

git clone https://github.com/jasonlcs/waferlock-robot-mcp-public.git
cd waferlock-robot-mcp-public

2. 編譯

npm install
npm run build

3. 直接執行 (本地調試)

node dist/cli.js \
  --api-url https://waferlock-robot-mcp-1177c207c107.herokuapp.com \
  --api-token <你的 API Token> \
  --mcp-token <選填：要求 MCP 客戶端提供的 Token>

  • --server-name：覆寫 MCP server 名稱

  • --server-version：覆寫 MCP server 版本

  • --mcp-token：若指定，MCP 客戶端必須提供同樣的 Token（透過環境變數或連線設定）

4. 透過 npx 執行 (Git 來源)

建立 tag 後（例如 git tag v1.0.0 && git push origin v1.0.0），即可提供使用者以下指令：

npx --yes github:jasonlcs/waferlock-robot-mcp-public#v1.0.0 \
  --api-url https://waferlock-robot-mcp-1177c207c107.herokuapp.com \
  --api-token <你的 API Token> \
  --mcp-token <選填>

5. ChatGPT Desktop 設定範例

{
  "mcpServers": {
    "waferlock-robot": {
      "command": "npx",
      "args": [
        "--yes",
        "github:jasonlcs/waferlock-robot-mcp-public#v1.0.0",
        "--api-url", "https://waferlock-robot-mcp-1177c207c107.herokuapp.com",
        "--api-token", "<你的 API Token>",
        "--mcp-token", "<選填>"
      ]
    }
  }
}

將 jasonlcs、waferlock-robot-mcp-1177c207c107、<你的 API Token> 改成實際值即可使用。

專案結構

waferlock-robot-mcp-public/
├── package.json
├── tsconfig.json
├── README.md
└── src
    ├── cli.ts
    └── services
        ├── manualApiProvider.ts
        ├── manualProvider.ts
        └── mcpService.ts

開發注意事項

  • 需要 Node.js 18+（提供原生 fetch）。

  • CLI 會優先讀取命令列參數（--api-url、--api-token 等），若未提供則回退至環境變數 API_URL、API_TOKEN、MCP_TOKEN。

  • 內建 dotenv，會載入 .env（選用）。

  • 專案使用 TypeScript；若想讓使用者省去編譯步驟，可在 repo 中同步 dist/，或加上 npm prepare 腳本在安裝時自動編譯。

後續步驟

  1. npm run build 確認 dist/ 生成。必要時 chmod +x dist/cli.js。

  2. git add . && git commit -m "Initial public CLI"，再 git push -u origin main。

  3. 建立版本 tag（例：git tag v1.0.0 && git push origin v1.0.0）。

  4. 在主專案或文件提到此 CLI，指引用法（npx github:... 或 node dist/cli.js ...）。

授權

ISC License

Available Tools

3 tools
get_manual_infoB

Get information about a specific manual by ID

Inputs: fileId (required).

list_manualsB

List all uploaded Waferlock product manuals

Inputs: no named parameters.

search_manualsC

Search for manuals by filename

Inputs: query (required).

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Deploy Server
    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    TDQS

    B3.2/5.0

    Scored across 3 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose: get_manual_info retrieves details for a specific manual, list_manuals shows all manuals, and search_manuals finds manuals by filename. There is no overlap or ambiguity in their functions.

    Naming Consistency5/5

    All tools follow a consistent verb_noun pattern (get_manual_info, list_manuals, search_manuals) with clear, descriptive names. There are no deviations in naming conventions.

    Tool Count3/5

    With only 3 tools, the set feels thin for a product manual management system. While the tools cover basic retrieval and listing, it lacks operations like upload, update, or delete, which are typical for such a domain.

    Completeness2/5

    The tool set is significantly incomplete for managing Waferlock product manuals. It only supports read operations (get, list, search), missing essential CRUD functions such as create/upload, update, and delete, which are necessary for full lifecycle coverage.

    Resources

    Related MCP Connectors