Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@hsh-erp-mcp Create an MCP tool for exporting purchase orders via SRM"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
hsh-erp monorepo 的 MCP（Model Context Protocol）服务器子模块脚手架。
当前实现：一个 echo hello world 工具，零外部依赖。后续在此基础上扩展业务工具。
tools/srm/：从 hsh-erp 旧 e2e/scripts/（与测试同仓）迁来的 SRM 浏览器自动化脚本（采购订单导出 / 历史价批量下载 / 应标项目下载）。后续会把 srm-export / srm-historical-price / srm-bid-download 包装成 MCP tools（每个工具就是一个 server.registerTool(...)），当前仍以独立 tsx + docker-run.sh 在 Playwright Docker 镜像内跑。
npm install # 安装依赖（首次）
npm start # 启动 stdio 服务器（= tsx src/index.ts）
npm run inspect # 用 @modelcontextprotocol/inspector 在浏览器调试
# SRM 浏览器自动化（独立子项目）
npm run srm:export -- 6200039492 # 采购订单导出
npm run srm:export:headed -- 6200039492 # headed 调试Related MCP server: claw-mcp-toolkit
mcp/
├── package.json # 依赖与脚本（type=module, tsx 直接执行 TS）；npm scripts 含 srm:export 系列
├── tsconfig.json # ES2022 + NodeNext
├── src/
│ └── index.ts # 服务器入口 + createServer() 工厂 + registerTool()
├── tools/
│ └── srm/ # 从 e2e/scripts/ 迁来（详见 tools/srm/README.md）
│ ├── srm-export.ts # 采购订单 Excel
│ ├── srm-historical-price.ts # 历史价批量下载
│ ├── srm-bid-download.ts # 应标项目批量下载
│ ├── env.ts # 凭据读取
│ ├── docker-run.sh # Playwright Docker 一键封装
│ ├── package.json # 子项目依赖（playwright + tsx）
│ ├── tsconfig.json
│ ├── .env.example # SRM_USERNAME / SRM_PASSWORD 模板
│ └── README.md
└── ...在 src/index.ts 的 createServer() 里继续调用 server.registerTool(...)：
server.registerTool(
'my-tool',
{
description: '...',
inputSchema: z.object({ ... }), // zod v4 单 schema，handler 类型由 SDK 推导
},
async (args) => ({
content: [{ type: 'text', text: '...' }],
})
);每个工具的 inputSchema 必须是单一 zod schema，handler 返回值结构：
{ content: [{ type: 'text', text: '...' }], isError?: boolean }npm run inspect 启动 inspector 后浏览器操作：
Connect
Tools 标签
选工具，填参数，Call
也可在 inspector 里观察 initialize / tools/list / tools/call 的 JSON-RPC 消息体，确认 stdout 通道无杂输出。
stdout 是 JSON-RPC 协议流——禁止 console.log / process.stdout.write，调试只准走 console.error。
zod 锁 v4，import 用 zod/v4，不要 混 v3。
子模块内提交 → 推送子模块远端 → 根仓库 git add mcp 提交新 SHA。
Echo back the input message. Hello-world scaffold for hsh-erp MCP server.
| Name | Required | Description | Default |
|---|---|---|---|
| message | Yes | Text to echo back |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the burden of behavioral disclosure. 'Echo back the input message' clearly conveys the operation and its lack of side effects in this trivial case, and 'Hello-world scaffold' sets expectations of simplicity.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Two short sentences, no wasted words. The core behavior is front-loaded and the scaffold context is provided in a single additional clause.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a one-parameter echo tool with no output schema and no sibling tools, the description is fully sufficient. An agent can correctly select and invoke this tool based on the provided information alone.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%: the message parameter is already described as 'Text to echo back.' The description adds no additional parameter semantics, but none are needed given the schema already fully documents the only parameter.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Description states a specific verb and resource: 'Echo back the input message.' It is immediately clear what the tool does and what data it operates on. The 'Hello-world scaffold' phrase adds useful context about its role.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies this is a test/demo scaffold for the hsh-erp MCP server, which suggests when it might be used. However, it does not explicitly state when to use it over alternatives or when not to use it, though no siblings exist to differentiate from.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
With only one tool, there is no possibility of confusing it with other tools. The echo tool is clearly distinct simply by being the only tool available.
A single tool name cannot exhibit inconsistency. 'echo' is a simple, valid name, and the lack of other tools means no conflicting naming patterns exist.
An ERP server with a single echo/hello-world tool is an extreme mismatch. The tool is trivial and entirely unrelated to ERP functionality, making the count grossly inappropriate for the stated domain.
The server claims to be an ERP MCP server but only provides an echo tool. There are no operations for any ERP entities or workflows, resulting in a severely incomplete surface.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Search multi-merchant supply, checkout, and track orders via MCP.
Access all your MCPBundles tools in one place. Unified hub for all enabled bundles.
Create, schedule, and publish social posts, manage accounts, and read analytics as MCP tools.
Build, deploy and manage MaShop e-commerce projects from Claude, Cursor or any MCP client.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/kurisu-004/hsh-erp-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
