💬 Auto-MSG-Line-Notify

Auto-MSG-Line-Notify is the easiest on-ramp in the whole auto-messaging pack, and it's a little unfair how much simpler it is than its siblings. Where Telegram needs a bot token plus a chat ID and Discord needs a developer portal account plus channel IDs, LINE Notify needs exactly one thing: an access token. Paste it in, flip one switch, and your renders start landing in your LINE app.

This is the node to try first if you've never used the pack. You'd reach for it any time you want ComfyUI to ping your phone when a long batch finishes - kick off 30 images, leave the desk, come back to the results waiting in chat. The author is Taiwan-based, and LINE is the messaging app there the way Telegram is everywhere else; that's why a small pack by a Taiwanese dev ships a LINE option when most Western ones skip it.

How it works

Same shared machinery as the rest of the pack: the node converts the incoming IMAGE tensor to a PNG and POSTs it to https://notify-api.line.me/api/notify with an Authorization: Bearer <token> header, uploading the image as a multipart file. Your prompt JSON (truncated to 500 characters) is appended to the message. No bot, no channel, no webhook server - ComfyUI is the sender, LINE Notify is just an endpoint.

The two inputs that matter are im_line_notify_token and enable_im_line_notify. Add im_line_notify_msg_header if you want a custom label, feed trigger_append_image your VAE-decoded output, and you're done. The msg_png_info_enabled / msg_pos_prompt_enabled / msg_neg_prompt_enabled toggles are the same accepted-but-inert booleans the whole pack shares - the prompt is always included, so don't fuss over them.

Outputs are the standard pass-through pair (output-original-input, output-original-input-IMAGE) plus three STRING outputs. 🌀output-text-LINE holds the API's JSON response, which is your debug readout when nothing shows up.

Getting the token

1. Go to notify-bot.line.me and log in with LINE.
2. Generate a token - you can scope it to your own chat or to a group.
3. Copy it into im_line_notify_token.

Per the README, it's free for 1,000 requests per token, and each account can hold up to 100 tokens. For personal use that quota is a rounding error; only an unattended overnight loop will ever get near it.

Install

ComfyUI Manager: search "ComfyUI-decadetw-auto-messaging-realtime". Or:

cd ComfyUI/custom_nodes
git clone https://github.com/xlinx/ComfyUI-decadetw-auto-messaging-realtime

Restart ComfyUI. No requirements.txt and no model downloads - just requests and Pillow, both already in a stock ComfyUI.

Common issues

• Nothing arrives? Check the 🌀output-text-LINE output. A 401 means the token is wrong or the service rejected it; a 429 means you hit the request limit.
• The token is a plaintext widget in the workflow JSON. Scrub it before sharing a workflow - it's tied to your LINE account, and anyone holding it can post to your Notify targets.
• One honest heads-up: LINE has announced plans to retire Notify, so treat this as the quick-win channel for personal use rather than something to build a long-term pipeline on. For that, the Telegram node in this same pack is the more future-proof bet.