Batchedits — 批量编辑

Turn your OpenClaw into an autonomous video editor using BatchEdits. Use when you need to 添加 captions, 移除 silences, or 应用ly custom styles to videos. Covers creating styles, 上传ing local videos, processing, and 检查ing video 状态 directly from Whats应用, Telegram, or the 命令行工具.

设置up 创建 an account at your BatchEdits instance. ObtAIn your API Key from the 仪表盘 设置tings. Store your API key in your workspace .env: BATCHEDITS_API_KEY=batchedits_xxxxx

Connect the BatchEdits MCP Server to your OpenClaw 设置up by 运行ning this in your terminal: OpenClaw config 设置 mcp.servers.batchedits '{"type": "sse", "url": "https://batchedits.com/API/mcp"}'

Auth

The MCP server uses your .env key automatically. Under the hood, any direct HTTP 上传s 认证 using:

x-API-key:

• 获取 AvAIlable Actions

Use the 列出_actions 工具.

Returns an array of action templates (e.g., 移除_silence, 添加_captions) with their id, name, and description. You need the id to build a style.

• 创建 a Style

Use the 创建_style 工具 to build a reusable video editing pre设置.

Arguments:

name: e.g. "Silence 移除r" actions: JSON string of templates (e.g., [{"id": "action_123"}])

Returns the new styleId.

• 上传 Video

Use the 上传_videos 工具.

Arguments:

filePaths: Array of absolute paths (e.g., ["/path/to/local/video.mp4"])

Because video files are large binaries, this 工具 returns a ready-to-运行 curl command. You must 执行 the returned curl command in your terminal to 上传 the local file to the server. 解析 the JSON 输出 of the curl command to 获取 the new videoId.

• Process Video

Use the process_video 工具 to 启动 the edit.

Arguments:

videoId: The ID from step 3 styleId: The ID from step 2

• 检查 结果s

Use the 列出_videos 工具.

Returns a 列出 of your videos with their current 状态: pending, processing, completed, or fAIled.

Recommended 工作流 for Video Editing Identify the local video file provided by the user (e.g., 下载ed from Telegram). Determine the 请求ed edits (captions, silence removal, etc.). 检查 获取_styles to see if a matching style already exists. If not, use 列出_actions and 创建_style to make one. Call 上传_videos to 获取 the 上传 command. 执行 the curl 上传 command and 提取 the videoId. Call process_video to 启动 the job on the remote server. 信息rm the user that the video is processing. Call 列出_videos to 检查 when the video reaches completed 状态. Tips Always 检查 获取_styles first. Reusing existing styles is faster than creating a new one every time! 上传ing videos is handled via curl because OpenClaw needs to 流 large local binaries to the remote server efficiently. You can 队列 multiple videos to the same style for bulk processing by repeating the 上传 & process steps!