setup

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

MeiGen Plugin Setup

MeiGen插件设置

You are guiding the user through configuring the MeiGen plugin for image generation. Follow this flow step by step.

你将引导用户完成用于图片生成的MeiGen插件配置，请按以下步骤逐步操作。

Step 1: Welcome

步骤1：欢迎

First, check if a config file already exists:

bash

cat ~/.config/meigen/config.json 2>/dev/null

Also check for existing ComfyUI workflows:

bash

ls ~/.config/meigen/workflows/*.json 2>/dev/null

If config exists, show the current configuration (mask API keys: show first 10 chars + "...") and any saved workflows. Ask if they want to reconfigure.

If no config exists, present this introduction:

MeiGen Plugin Configuration

This is optional. Without configuration, you can still use free features:

Search gallery for inspiration and prompts

Enhance simple ideas into professional prompts

Browse available AI models

Configuring a provider unlocks image generation.

Then proceed to Step 2.

首先，检查配置文件是否已存在：

bash

cat ~/.config/meigen/config.json 2>/dev/null

同时检查是否存在已有的ComfyUI工作流：

bash

ls ~/.config/meigen/workflows/*.json 2>/dev/null

如果配置已存在，显示当前配置（隐藏API密钥：显示前10个字符+“...”）以及所有已保存的工作流。询问用户是否需要重新配置。

如果配置不存在，展示以下介绍内容：

MeiGen插件配置

此配置为可选操作。未配置时，你仍可使用免费功能：

搜索灵感图库及提示词

将简单想法优化为专业提示词

浏览可用AI模型

配置提供商后即可解锁图片生成功能。

之后进入步骤2。

Step 2: Choose Provider

步骤2：选择提供商

Present these options to the user:

为用户提供以下选项：

Option A: MeiGen Platform (Recommended)

选项A：MeiGen平台（推荐）

Supports Nanobanana Pro, GPT image 1.5, Seedream 4.5 and more
Reference image support for style transfer
No additional accounts needed — just get a token from meigen.ai

支持Nanobanana Pro、GPT image 1.5、Seedream 4.5等更多模型
支持参考图片进行风格迁移
无需额外注册账号——只需从meigen.ai获取令牌

Option B: ComfyUI (Local)

选项B：ComfyUI（本地）

Use your local ComfyUI installation for image generation
Full control over models, samplers, and workflow settings
No cloud API needed — runs entirely on your machine
Import your own workflow from ComfyUI

使用本地ComfyUI安装环境进行图片生成
完全控制模型、采样器及工作流设置
无需云API——完全在本地设备运行
可从ComfyUI导入自定义工作流

Option C: Custom OpenAI-Compatible API

选项C：自定义OpenAI兼容API

Use your own OpenAI, Together AI, Fireworks AI, or any OpenAI-compatible service
Bring your own API key and billing
Supports any model that uses the OpenAI
```
/v1/images/generations
```
endpoint

使用自有OpenAI、Together AI、Fireworks AI或任何OpenAI兼容服务
自带API密钥及计费方式
支持所有使用OpenAI
```
/v1/images/generations
```
接口的模型

Option D: Import from curl Example

选项D：从curl示例导入

Already have a working curl command from your API provider's docs? Paste it directly!
We'll automatically extract the API key, base URL, and model name

已有API提供商文档中的可用curl命令？直接粘贴即可！
系统将自动提取API密钥、基础URL及模型名称

Option E: Skip image generation for now

选项E：暂时跳过图片生成配置

Free features still available (inspiration search, prompt enhancement, model listing)
You can run
```
/meigen:setup
```
anytime later to enable image generation

If user chooses Skip, say goodbye and exit. Otherwise continue to the appropriate Step 3.

仍可使用免费功能（灵感搜索、提示词优化、模型列表浏览）
之后随时运行
```
/meigen:setup
```
即可启用图片生成功能

如果用户选择跳过，告知用户并退出流程。否则进入对应步骤3。

Step 3A: MeiGen Platform Setup

步骤3A：MeiGen平台设置

Ask the user:

Do you already have a MeiGen API token, or do you need to create one?

询问用户：

你是否已拥有MeiGen API令牌，还是需要创建新令牌？

If they need to create one:

如果用户需要创建令牌：

Provide these instructions:

Go to https://www.meigen.ai
Sign in or create an account
Navigate to Settings (click your avatar) → API Keys
Click Create API Key, give it a name
Copy the token (starts with
```
meigen_sk_
```
)

Then ask them to paste the token.

提供以下操作说明：

访问https://www.meigen.ai
登录或创建账号
进入设置（点击头像）→ API密钥
点击创建API密钥，为其命名
复制令牌（以
```
meigen_sk_
```
开头）

之后请用户粘贴该令牌。

If they already have one:

如果用户已拥有令牌：

Ask them to paste their

meigen_sk_...

token.

请用户粘贴其

meigen_sk_...

格式的令牌。

Validate the token:

验证令牌：

Must start with
```
meigen_sk_
```
Must be at least 30 characters long

If valid, proceed to Step 4 with this config:

json

{
  "meigenApiToken": "<the token>"
}

必须以
```
meigen_sk_
```
开头
长度至少为30个字符

如果验证通过，使用以下配置进入步骤4：

json

{
  "meigenApiToken": "<令牌内容>"
}

Step 3B: ComfyUI (Local) Setup

步骤3B：ComfyUI（本地）设置

3B-1: Check Connection

3B-1：检查连接

Ask the user for their ComfyUI server URL:

What is your ComfyUI server URL? (default:
http://localhost:8188
)

Test the connection:

bash

curl -s <URL>/system_stats | head -c 200

Success: Show confirmation and continue to 3B-2
Failure: Tell the user:
Cannot connect to ComfyUI at
```
<URL>
```
. Please make sure:
1. ComfyUI is running (start it with
```
python main.py
```
  or your launcher)
2. The URL and port are correct
3. No firewall is blocking the connection

询问用户ComfyUI服务器地址：

你的ComfyUI服务器地址是什么？（默认：
http://localhost:8188
）

测试连接：

bash

curl -s <URL>/system_stats | head -c 200

连接成功：显示确认信息并进入3B-2
连接失败：告知用户：
无法连接到
```
<URL>
```
地址的ComfyUI。请确保：
1. ComfyUI已启动（使用
```
python main.py
```
  或启动器运行）
2. 地址及端口正确
3. 无防火墙阻止连接

3B-2: Import Workflow

3B-2：导入工作流

Explain the workflow export process to the user:

To use ComfyUI with this plugin, you need to export a workflow in API format:
Open ComfyUI in your browser (usually
http://localhost:8188
)
Load or create your preferred workflow

Click ⚙️ Settings → enable "Enable Dev mode options"

Click the "Save (API Format)" button that appears
Save the downloaded
.json
file somewhere convenient

Then ask them to provide the file path:

Please provide the path to your exported workflow JSON file: Example:
~/Downloads/workflow_api.json

Use the

comfyui_workflow

tool with action

import

to import the workflow:

Ask for a short name for the workflow (e.g., "txt2img", "anime", "realistic")
Call:
```
comfyui_workflow import
```
with
```
filePath
```
and
```
name
```
Show the detected nodes and parameters to the user for confirmation

If the import succeeds, ask if they want to import additional workflows. If yes, repeat this step.

向用户说明工作流导出流程：

要在本插件中使用ComfyUI，你需要以API格式导出工作流：
在浏览器中打开ComfyUI（通常为
http://localhost:8188
）
加载或创建你偏好的工作流

点击**⚙️ 设置** → 启用**“启用开发者模式选项”**

点击出现的**“保存（API格式）”**按钮
将下载的
.json
文件保存至方便访问的位置

之后请用户提供文件路径：

请提供你导出的工作流JSON文件路径：示例：
~/Downloads/workflow_api.json

使用

comfyui_workflow

工具并指定

import

动作来导入工作流：

请用户为工作流设置简短名称（例如：“txt2img”、“anime”、“realistic”）
调用：
```
comfyui_workflow import
```
并传入
```
filePath
```
和
```
name
```
参数
向用户展示检测到的节点及参数以确认

如果导入成功，询问用户是否需要导入更多工作流。如果是，重复此步骤。

3B-3: Save Configuration

3B-3：保存配置

Build the config JSON:

json

{
  "comfyuiUrl": "<the URL, omit if http://localhost:8188>",
  "comfyuiDefaultWorkflow": "<the first imported workflow name>"
}

Proceed to Step 4 to save. The workflow files are already saved by the import step.

构建配置JSON：

json

{
  "comfyuiUrl": "<服务器地址，默认地址可省略>",
  "comfyuiDefaultWorkflow": "<首个导入的工作流名称>"
}

进入步骤4保存配置。工作流文件已在导入步骤中完成保存。

Step 3C: Custom OpenAI-Compatible API Setup

步骤3C：自定义OpenAI兼容API设置

Collect the following information. Present common presets first for convenience:

收集以下信息。为方便操作，先展示常用预设：

Quick Presets

快速预设

Service	Base URL	Default Model
OpenAI	`https://api.openai.com` (default)	`gpt-image-1.5`
Together AI	`https://api.together.xyz/v1`	(check their docs)
Fireworks AI	`https://api.fireworks.ai/inference/v1`	(check their docs)

Ask the user to either pick a preset or provide custom values.

服务	基础URL	默认模型
OpenAI	`https://api.openai.com` （默认）	`gpt-image-1.5`
Together AI	`https://api.together.xyz/v1`	（请查阅其官方文档）
Fireworks AI	`https://api.fireworks.ai/inference/v1`	（请查阅其官方文档）

询问用户选择预设或提供自定义参数。

Required Fields

必填字段

API Key (required): Their API key for the service
- Example:
```
sk-...
```
  for OpenAI
Base URL (optional): API endpoint URL
- Default:
```
https://api.openai.com
```
- Only needed if using a non-OpenAI service
Model Name (optional): Which model to use
- Default:
```
gpt-image-1.5
```
- Different services use different model names

API密钥（必填）：对应服务的API密钥
- 示例：OpenAI的密钥格式为
```
sk-...
```
基础URL（可选）：API接口地址
- 默认：
```
https://api.openai.com
```
- 仅在使用非OpenAI服务时需要填写
模型名称（可选）：要使用的模型
- 默认：
```
gpt-image-1.5
```
- 不同服务使用的模型名称不同

Optional: Test the connection

可选：测试连接

After collecting the info, suggest testing with curl:

bash

curl -s <BASE_URL>/v1/models \
  -H "Authorization: Bearer <API_KEY>" | head -c 500

This helps catch invalid keys or wrong URLs before saving.

Proceed to Step 4 with config from the collected fields (see bottom of this section for format).

Only include fields that differ from defaults. Omit

openaiBaseUrl

if it's

https://api.openai.com

, omit

openaiModel

if it's

gpt-image-1.5

收集信息后，建议使用curl测试连接：

bash

curl -s <BASE_URL>/v1/models \
  -H "Authorization: Bearer <API_KEY>" | head -c 500

这有助于在保存配置前发现无效密钥或错误地址。

使用收集到的字段构建配置，进入步骤4（格式见本节末尾）。

仅保留与默认值不同的字段。如果

openaiBaseUrl

为

https://api.openai.com

则省略，如果

openaiModel

为

gpt-image-1.5

则省略。

Step 3D: Import from curl Example

步骤3D：从curl示例导入

Ask the user to paste their curl command. Common formats they might paste:

Format 1: Image generation endpoint

bash

curl https://api.openai.com/v1/images/generations \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{"model": "gpt-image-1.5", "prompt": "a cat", "n": 1, "size": "1024x1024"}'

Format 2: Simple model list test

bash

curl https://api.together.xyz/v1/models \
  -H "Authorization: Bearer xxx"

Format 3: With -u flag or other auth styles

bash

curl -u :sk-xxx https://api.fireworks.ai/inference/v1/images/generations \
  -d '{"model": "accounts/fireworks/models/flux", "prompt": "a cat"}'

请用户粘贴其curl命令。用户可能粘贴的常见格式如下：

格式1：图片生成接口

bash

curl https://api.openai.com/v1/images/generations \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{"model": "gpt-image-1.5", "prompt": "a cat", "n": 1, "size": "1024x1024"}'

格式2：简单模型列表测试

bash

curl https://api.together.xyz/v1/models \
  -H "Authorization: Bearer xxx"

格式3：带-u标志或其他认证方式

bash

curl -u :sk-xxx https://api.fireworks.ai/inference/v1/images/generations \
  -d '{"model": "accounts/fireworks/models/flux", "prompt": "a cat"}'

Parse the curl command and extract:

解析curl命令并提取信息：

Base URL: The URL hostname + base path (strip

/v1/images/generations

/v1/models

, etc.)

https://api.openai.com/v1/images/generations

→

https://api.openai.com

https://api.together.xyz/v1/models

→

https://api.together.xyz/v1

If URL ends with
```
/v1/...
```
, keep the
```
/v1
```
part only if it's NOT
```
api.openai.com
```

API Key: From

Authorization: Bearer <key>

header, or

-u :<key>

flag, or

--header

variants

Model: From the JSON request body
```
"model": "<value>"
```
if present

基础URL：URL主机名+基础路径（移除

/v1/images/generations

、

/v1/models

等后缀）

https://api.openai.com/v1/images/generations

→

https://api.openai.com

https://api.together.xyz/v1/models

→

https://api.together.xyz/v1

如果URL以
```
/v1/...
```
结尾，仅当地址不是
```
api.openai.com
```
时保留
```
/v1
```
部分

API密钥：从
```
Authorization: Bearer <密钥>
```
请求头、
```
-u :<密钥>
```
标志或
```
--header
```
变体中提取
模型：如果JSON请求体中存在
```
"model": "<值>"
```
，则提取对应值

Show parsed results for confirmation:

展示解析结果供用户确认：

I extracted the following from your curl command:
API Key:
sk-xxx...
(first 10 chars)
Base URL:
https://api.together.xyz/v1
Model:
black-forest-labs/FLUX.1-schnell
Does this look correct?

If user confirms, proceed to Step 4. If not, let them correct individual fields.

已从你的curl命令中提取以下信息：
API密钥：
sk-xxx...
（前10个字符）
基础URL：
https://api.together.xyz/v1
模型：
black-forest-labs/FLUX.1-schnell
以上信息是否正确？

如果用户确认，进入步骤4。如果不正确，允许用户修改单个字段。

Step 4: Save Configuration

步骤4：保存配置

Build the config JSON based on the chosen provider:

For MeiGen:

json

{
  "meigenApiToken": "<the token>"
}

For ComfyUI:

json

{
  "comfyuiUrl": "<url, omit if default>",
  "comfyuiDefaultWorkflow": "<workflow name>"
}

For OpenAI-compatible (manual or curl import):

json

{
  "openaiApiKey": "<the key>",
  "openaiBaseUrl": "<base url, omit if default>",
  "openaiModel": "<model, omit if default>"
}

Create the config directory and write the file:

bash

mkdir -p ~/.config/meigen

Then use the Write tool to write the JSON config to

~/.config/meigen/config.json

Important: If the user already has a config file with other providers configured, merge the new config into the existing one rather than overwriting. For example, a user might have both MeiGen and ComfyUI configured.

After writing, set permissions:

bash

chmod 600 ~/.config/meigen/config.json

根据用户选择的提供商构建配置JSON：

MeiGen平台配置：

json

{
  "meigenApiToken": "<令牌内容>"
}

ComfyUI配置：

json

{
  "comfyuiUrl": "<服务器地址，默认地址可省略>",
  "comfyuiDefaultWorkflow": "<工作流名称>"
}

OpenAI兼容API配置（手动或curl导入）：

json

{
  "openaiApiKey": "<密钥内容>",
  "openaiBaseUrl": "<基础URL，默认地址可省略>",
  "openaiModel": "<模型名称，默认模型可省略>"
}

创建配置目录并写入文件：

bash

mkdir -p ~/.config/meigen

然后使用写入工具将JSON配置写入

~/.config/meigen/config.json

。

重要提示：如果用户已存在包含其他提供商配置的文件，请将新配置合并到现有文件中，而非覆盖。例如，用户可能同时配置了MeiGen和ComfyUI。

写入完成后，设置文件权限：

bash

chmod 600 ~/.config/meigen/config.json

Step 5: Completion

步骤5：配置完成

Tell the user:

Configuration saved! To activate the new settings, please start a new Claude Code session (close and reopen, or open a new terminal tab).

After restarting, you can:
Use
generate_image
to create AI images
Run
list_models
to see available models and workflows
Try: "Generate a beautiful sunset over mountains"
You can run
/meigen:setup
again anytime to change your configuration.

For ComfyUI users, additionally mention:

ComfyUI tips:
Use
comfyui_workflow list
to see your saved workflows
Use
comfyui_workflow view
to see adjustable parameters (steps, CFG, sampler, etc.)
Ask me to change any workflow parameter — e.g., "increase steps to 30" or "switch sampler to dpmpp_2m"
You can import more workflows anytime with
comfyui_workflow import

告知用户：

配置已保存！请重启Claude Code会话（关闭并重新打开，或打开新终端标签页）以激活新设置。

重启后，你可以：
使用
generate_image
命令生成AI图片
运行
list_models
查看可用模型及工作流
尝试输入：“生成一幅山脉上的绝美日落图”
你可随时再次运行
/meigen:setup
修改配置。

针对ComfyUI用户，额外补充：

ComfyUI使用提示：
使用
comfyui_workflow list
查看已保存的工作流
使用
comfyui_workflow view
查看可调整参数（步数、CFG、采样器等）
可要求修改任意工作流参数——例如：“将步数增加至30”或“切换采样器为dpmpp_2m”
你可随时使用
comfyui_workflow import
导入更多工作流