Skip to main content
Glama
ZCode Image Search MCP

ZCode Image Search MCP

by umineko987
README.md
# ZCode Image Search MCP

在 Claude Code、Codex 或 Pi 中调用 ZCode 的 `search_image`:搜索**现有图片**,不生成图片。服务通过本地 stdio 运行,请求转发到 ZCode 搜图接口。

## 使用前

需要 Node.js ≥ 20,以及可使用搜图服务的 Z.ai 或 BigModel 账号。Pi 本身需要 Node.js ≥ 22.19。

在**运行客户端的同一台机器、同一用户环境**下完成以下任一项:

- 已登录 ZCode:服务启动时会尝试读取并同步本地登录凭据;或
- 在终端单独登录(将 `zai` 换成 `bigmodel` 可使用 BigModel):

```bash
npx -y @umineko987/zcode-image-search-mcp@latest login zai
```

## 接入客户端

### Claude Code

```bash
claude mcp add zcode-image-search -- npx -y @umineko987/zcode-image-search-mcp@latest
```

默认仅对当前项目生效;如果希望所有项目可用,在 `zcode-image-search` 前加 `--scope user`。

### Codex

```bash
codex mcp add zcode-image-search -- npx -y @umineko987/zcode-image-search-mcp@latest
```

### Pi

Pi 需要先安装第三方 [pi-mcp-adapter](https://github.com/nicobailon/pi-mcp-adapter):

```bash
pi install npm:pi-mcp-adapter
```

重启 Pi,并在 `~/.config/mcp/mcp.json` 中加入以下配置(如已有配置,将 `zcode-image-search` 合并到现有 `mcpServers`):

```json
{
  "mcpServers": {
    "zcode-image-search": {
      "command": "npx",
      "args": ["-y", "@umineko987/zcode-image-search-mcp@latest"]
    }
  }
}
```

在 Pi 中运行 `/mcp` 可查看连接状态。

## 使用

让客户端调用 `search_image`,传入 `query`(关键词)、`count`(数量)、`gl`(地区代码,如 `us`)、`rank`(是否排序);这四项均为必填。示例:“用 `search_image` 搜索布偶猫图片,`count=3`、`gl=us`、`rank=true`。”

TDQS

B3.4/5.0

Scored across 1 tool

Disambiguation5/5

There is only one tool, so there is no possibility of confusion or overlapping purpose with another tool.

Naming Consistency5/5

The single tool name 'search_image' follows a clear verb_noun pattern, and with only one tool there is no inconsistency to evaluate.

Tool Count3/5

A single tool is borderline thin for an image search server; while search is the core operation, additional tools for filtering, pagination, or image details could reasonably be expected.

Completeness4/5

The tool covers the primary search operation, but there are minor potential gaps such as result pagination, filtering, or retrieving image details that would enhance completeness.

Maintenance

ActivitySlowing
ResponsivenessNo recent community issues