Skip to main content
Glama
amazing1102

doubao-tools

by amazing1102
OverviewSchemaSourceRelated ServersDiscussionsAdmin
JavaScript
Local

doubao-tools

一个给 Claude Code + DeepSeek 工作流使用的 MCP 视觉分析工具。

当你主要用 Claude Code 接入 DeepSeek/其他模型写代码时，DeepSeek 侧可能不方便直接查看本地图片、截图或 UI 截图。这个 MCP 服务把本地图片交给豆包视觉模型分析，再把分析结果返回给 Claude Code，方便后续定位界面问题、阅读截图内容或给出修改建议。

适合场景

  • Claude Code 接 DeepSeek 写代码，但需要分析本地截图

  • 让 AI 查看 UI 截图，指出布局、文字、交互、可访问性问题

  • 根据图片内容生成修改建议，再回到代码里处理

  • 不想把截图手动上传到多个网页工具

Related MCP server: Doubao Vision MCP Server

功能

  • 提供 doubao_vision_analyze MCP 工具

  • 读取本地图片文件并转成 data URL

  • 调用火山方舟兼容 OpenAI SDK 的 Chat Completions 接口

  • 支持自定义 prompt，适合 UI 截图检查、图片描述和视觉问题排查

环境要求

  • Node.js 18 或更高版本

  • 火山方舟 API Key

  • 可用的豆包视觉模型 ID（如 doubao-seed-2-0-mini-260428，后面的模型 ID填写时要有模型的版本号）

  • Claude Code 或其他支持 MCP 的客户端

获取凭证

API Key

  1. 访问火山方舟控制台

  2. 在「模型推理」→「API Key 管理」中创建 API Key

  3. 复制 Key 备用

视觉模型 ID

在火山方舟控制台的「在线推理」页面，找到你可用的豆包视觉模型（如图像理解、UI 截图分析等模型），复制其模型 ID（格式通常为 doubao-seed-2-0-mini-260428 或类似）。

安装

# 克隆仓库
git clone https://github.com/amazing1102/doubao-tools-MCP.git
cd doubao-tools

# 安装依赖
npm install

配置环境变量

复制示例文件并填入自己的配置：

cp .env.example .env

编辑 .env 文件，填入实际值：

ARK_API_KEY=your_ark_api_key
ARK_BASE_URL=https://ark.cn-beijing.volces.com/api/v3
DOUBAO_VISION_MODEL=your_doubao_vision_model

Windows 用户也可以把这些变量设置到系统环境变量中，然后通过 run-server.cmd 启动，无需 .env 文件。

Claude Code MCP 配置

编辑 Claude Code 的 MCP 配置文件（通常位于 ~/.claude/claude_desktop_config.json 或通过 /mcp 命令管理），添加以下内容：

{
  "mcpServers": {
    "doubao-tools": {
      "command": "node",
      "args": ["/path/to/doubao-tools/server.js"],
      "env": {
        "ARK_API_KEY": "your_ark_api_key",
        "ARK_BASE_URL": "https://ark.cn-beijing.volces.com/api/v3",
        "DOUBAO_VISION_MODEL": "your_doubao_vision_model"
      }
    }
  }
}

注意：将 /path/to/doubao-tools/server.js 替换为本机仓库的实际路径。

  • Windows 示例: C:/Users/你的用户名/doubao-tools/server.js

  • macOS/Linux 示例: /home/你的用户名/doubao-tools/server.js

如果已经把环境变量设置到系统中，MCP 配置可以简化为：

{
  "mcpServers": {
    "doubao-tools": {
      "command": "node",
      "args": ["/path/to/doubao-tools/server.js"]
    }
  }
}

Windows 启动脚本

如果你把环境变量保存到了 Windows 用户环境变量里，可以使用启动脚本：

run-server.cmd

该脚本会从用户环境变量中读取 ARK_API_KEY、ARK_BASE_URL、DOUBAO_VISION_MODEL 并启动服务。

工具参数

doubao_vision_analyze

参数

类型

必填

默认值

说明

image_path

string

是

-

本地图片的绝对路径

prompt

string

否

详细中文 prompt

分析图片时使用的提示词

支持的图片格式：PNG、JPG、JPEG、GIF、BMP、WebP、TIFF。

使用方式示例

在 Claude Code 中可以让模型调用这个工具，例如：

请用 doubao_vision_analyze 分析这张截图：C:\Users\me\Desktop\screenshot.png
重点看 UI 有没有文字溢出、布局错位和可访问性问题。

豆包视觉模型会先读图并返回分析结果，然后你可以继续让 Claude Code/DeepSeek 根据结果修改代码。

常见问题

工具在 Claude Code 中不可用

检查 MCP 配置是否正确加载，确认 command 为 node，args 路径正确指向 server.js。

认证失败 (401/403)

检查 ARK_API_KEY 是否正确配置，确认 API Key 未过期且拥有对应模型的调用权限。

模型未找到 (404)

检查 DOUBAO_VISION_MODEL 模型 ID 是否拼写正确，确认该模型已在火山方舟控制台开通。

网络连接失败

如果不在中国内地，可能需要确认能否正常访问 ark.cn-beijing.volces.com，必要时配置代理。

许可协议

ISC

Available Tools

1 tool
doubao_vision_analyzeC

Analyze a local image or screenshot with a Doubao vision model.

ParametersJSON Schema
NameRequiredDescriptionDefault
promptNoWhat to inspect in the image.请详细描述这张图片。如果这是 UI 截图，请指出可见文字、布局问题、交互问题、可访问性问题，以及可能需要修改的代码方向。请用清晰的中文分点回答。
image_pathYesAbsolute path to the local image file.

TDQS

C2.9/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description must disclose behavioral traits. It only states that analysis occurs, but does not mention if the tool is read-only, requires network access, has rate limits, or what the output format is. Significant behavioral gaps exist.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single sentence with no fluff, front-loading the purpose. However, it is overly concise and omits critical contextual information that a few extra sentences could provide.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The tool has no output schema and no annotations, so the description should cover what the tool returns and any important context. It does not mention return format, error behavior, or file size limits, leaving the agent with incomplete information.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% and the schema descriptions are clear ('What to inspect in the image', 'Absolute path to the local image file'). The tool description adds no additional meaning beyond the schema, so a baseline score of 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose4/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description clearly states the tool's action (analyze) and resource (local image or screenshot) using a specific model (Doubao vision). However, the verb 'analyze' is somewhat generic and could be more specific about the type of analysis.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

No guidance is provided on when to use this tool, prerequisites (e.g., supported image formats, file existence), or when not to use it. The absence of sibling tools reduces the need for alternative comparisons, but usage context is still missing.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

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

  1. 1 tool updatev1.0.0
    • First observeddoubao_vision_analyze
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

B3.1/5.0

Scored across 1 tool

Disambiguation5/5

Only one tool exists, so there is no possible confusion with other tools.

Naming Consistency5/5

The single tool name follows a clear verb_noun pattern (analyze vision) and is internally consistent.

Tool Count2/5

With only one tool, the server feels too limited for its name 'doubao-tools', which implies multiple utilities.

Completeness2/5

The tool only provides image analysis; missing common supporting operations like uploading, listing, or comparing, making the surface incomplete.

Maintenance

ActivityInactive
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers