Generate new image element for canvas

Generate new video element for canvas

* @param {string} content - The content of the post * @returns {{tags: string[], content: string}} - The tags of the post and the content without tag

* Get user data directory * @returns {string} - User data directory path

* Get user data directory * @returns {string} - User data directory path

To get the correct path for pyinstaller bundled application

To get the correct path for pyinstaller bundled application

Retrieve the stream task associated with the given session_id. Args: session_id (str): Unique identifier for the session. Retur

Handle an incoming chat request. Workflow: - Parse incoming chat data. - Optionally inject system prompt. - Save chat session an

* Check if ComfyUI is installed * @returns {boolean} - True if ComfyUI is installed

* 检查PNG文件头

List all comfy workflows

List all chat sessions

Apply or rollback migrations to reach target version

* 从ArrayBuffer中读取字符串

注册单个工具

Remove the stream task associated with the given session_id. Args: session_id (str): Unique identifier for the session.

* Send cancellation message to main window or parent process * @param {string} message - Cancellation message

* Send error message to main window or parent process * @param {string} error - Error message

Send WebSocket notification about video generation error

* 设置环境变量代理 * @param {string} url 代理 URL

* Stop ComfyUI process * @returns {Promise<{success: boolean, message?: string}>}

Build API URL for Volces Cloud

Build request headers

Build request headers

Build request headers

Build a Pydantic model named '<WorkflowName>Input' from workflow['inputs']. The `inputs` column is stored in DB as JSON text -> parse first.

Build request payload based on whether input images are provided

Build request payload for Volces API

Build task search URL

Build request URL for Replicate API

Build workflow based on model type

Calculate width and height based on aspect ratio

Calculate width and height based on aspect ratio and model

Create OpenAI client for Volces API

创建语言模型实例

Ensure the database directory exists

修复聊天历史中不完整的工具调用 根据LangGraph文档建议，移除没有对应ToolMessage的tool_calls 参考: https://langchain-ai.github.io/langgraph/troubleshooting/errors/INVALID_CHAT

Generate image using Jaaz API service calling OpenAI model Compatible with OpenAI image generation API Returns:

Generate Replicate format image

Get the appropriate model for the request

处理单个chunk

处理错误

处理消息类型的 chunk

处理工具调用参数流

处理工具调用

处理 values 类型的 chunk

Initialize the database with the current schema

检查 Jaaz API 是否已配置

Send HTTP request and handle response Returns: dict[str, Any]: Response data from Replicate API

Normalize agent name to be compatible with tool names.

Poll for image generation result

Poll task status until completion

Process magic generation in a separate async task. Args: messages: List of messages session_id: Session ID canvas_id

Process Replicate API response and save image Args: res: Response data from Replicate API Returns:

Process OpenAI response and save image Args: result: OpenAI response object error_prefix: Error message pref

Map simple param types to Python types.

Search for existing cloud task Args: prompt: The generation prompt Returns: Task data if found and

等待确认

Return an @tool function for the given workflow record.

* Calculate SHA256 hash of a file * @param {string} filePath - Path to the file * @returns {Promise<string>} - Promise resolving to SHA256 hash in h

* Cancel the current ComfyUI installation

Clean up resources

* 清除代理设置 * @returns {Object} 空的代理环境变量对象

Compress an image to be under the specified size limit.

确认工具调用

创建所有智能体 Args: model: 语言模型实例 registered_tools: 已注册的工具名称列表 system_prompt: 系统提示词 Returns:

直接创建异步客户端（需要手动关闭）

Create a new canvas

Create a new comfy workflow

Create a tool that can handoff control to the requested agent. Args: agent_name: The name of the agent to handoff control to, i.e.

创建云端魔法图像生成任务 Args: image_content: 图片内容（base64 或 URL） Returns: str: 任务 ID，失败时返回空字符串

创建云端 Midjourney 图像生成任务 Args: prompt: 图像生成提示词 model: 图像生成模型（默认为 midjourney） **kwargs: 其他参数（如 mode

Factory method: create provider instance

创建云端视频生成任务 Args: prompt: 视频生成提示词 model: 视频生成模型 resolution: 视频分辨率 duration: 视频时长（秒）