wechat-files

star 12

Send and receive files, images, and documents through WeChat. Use when the user asks you to send a file, create a document, share an image, or when the user sends you a file to process.

lengyehan By lengyehan schedule Updated 2/24/2026

name: wechat-files description: "Send and receive files, images, and documents through WeChat. Use when the user asks you to send a file, create a document, share an image, or when the user sends you a file to process." metadata: { "openclaw": { "emoji": "📁" },

}

WeChat File Exchange

You can send and receive files through WeChat using the MEDIA: convention.

Sending files to the user

To send any file to the user via WeChat, output a MEDIA: line in your reply with the absolute path:

MEDIA:/absolute/path/to/file.xlsx

OpenClaw automatically detects the file type and sends it appropriately:

  • Images (png, jpg, gif, webp) are sent as WeChat image messages.
  • All other files (PDF, Excel, zip, code, etc.) are sent as file attachments.

Workflow

  1. Create or locate the file on disk (use exec or write tools).
  2. Include MEDIA:/path/to/file on its own line in your reply text.
  3. You may include explanatory text alongside the MEDIA line.

Examples

Create and send an Excel file:

I've created the spreadsheet for you.
MEDIA:/Users/username/.openclaw/workspace/report.xlsx

Send an existing file the user asked for:

Here's the file you requested.
MEDIA:/Users/username/Documents/presentation.pdf

Send a generated image:

Here's the chart I generated.
MEDIA:/Users/username/.openclaw/workspace/chart.png

Important rules

  • Always use absolute paths (starting with /).
  • Save generated files to ~/.openclaw/workspace/ by default.
  • Do not read binary files with cat; use appropriate tools for each format.
  • One MEDIA: line per file; use multiple lines for multiple files.
  • Do not wrap the MEDIA line in code fences or quotes.

Receiving files from the user

When a user sends a file or image through WeChat:

  • Images: Automatically injected into your context. If the model supports vision, you can see and analyze the image directly.
  • Documents/files (PDF, Excel, zip, etc.): Downloaded and saved locally. The file path is provided in your context. Use exec or read tools to process the file contents.

The context message will indicate what was received, for example:

  • [用户发送了一张图片] — with the image visible if model supports vision
  • [用户发送了文件: report.pdf] (saved: ~/.openclaw/media/wechat/report.pdf) — use read or exec to process
Install via CLI
npx skills add https://github.com/lengyehan/openclaw-wechat --skill wechat-files
Repository Details
star Stars 12
call_split Forks 3
navigation Branch main
article Path SKILL.md
More from Creator