macos-screenshot-telegram_skill

This skill captures a macOS screenshot and sends it to Telegram via Bot API, bypassing the OpenClaw media tool bug.

Python

2.5k

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill macos-screenshot-telegram

_meta.json308 B
SKILL.md2.4 KB

Overview

This skill captures a screenshot on macOS and reliably delivers it to a Telegram chat via the Telegram Bot API. It bypasses OpenClaw's built-in message tool bug by copying the screenshot into the OpenClaw workspace and sending the file with curl. Use a valid Telegram bot token and the target chat ID tied to your OpenClaw profile.

How this skill works

The skill runs macOS's screencapture to produce a PNG file, copies the image into the OpenClaw workspace (required by OpenClaw security restrictions), and posts the file directly to Telegram using the bot token and the sendPhoto endpoint. This avoids the OpenClaw message tool media bug by using the Telegram HTTP API for reliable media delivery.

When to use it

User requests a screen capture to be shared to Telegram
Sending a macOS screenshot from an OpenClaw environment where the message tool fails
Automating remote support or diagnostics via Telegram with a specific profile
You need to ensure screenshots are sent only to allowed chat IDs

Best practices

Confirm the correct OpenClaw profile name (e.g., main, rescue) before running the flow
Store the Telegram bot token in the profile config (~/.openclaw-<profile>/openclaw.json) under telegram.botToken
Add allowed chat IDs in telegram.allowFrom to restrict recipients
Always copy the screenshot into the workspace (~/.openclaw/workspace-<profile>/) before sending
Use the Telegram sendPhoto API (curl) instead of the OpenClaw message tool to avoid missing media delivery

Example use cases

Capture a macOS screen during a remote troubleshooting session and send it to the engineer's Telegram chat
Automated script: take periodic screenshots and forward them to a secure Telegram group for monitoring
Emergency capture: a user triggers a screenshot via an OpenClaw profile and sends it to a rescue Telegram chat
Share a desktop view with a developer or ops channel when diagnosing UI issues

FAQ

You need the Telegram bot token in telegram.botToken and the allowed chat ID in telegram.allowFrom inside ~/.openclaw-<profile>/openclaw.json.

Why copy the screenshot into the OpenClaw workspace?

OpenClaw restricts file access to its workspace. Copying the image into ~/.openclaw/workspace-<profile>/ ensures the skill can read and attach the file for upload.

How does this avoid the OpenClaw message tool bug?

Instead of using the native message tool that fails to send media, the skill calls Telegram's sendPhoto API directly with curl, which reliably delivers the image.