Give your n8n AI agent full chat memory (Get AI Context)
One node that hands your LLM the recent conversation and a clean transcript, so an AI reply actually knows what was said instead of answering blind.
Help center
Feature requests
Tell us what to build next and vote on what matters to you. When an idea you backed ships, we email you.
Suggest an idea
Ideas are reviewed before appearing on the board.
Guided warm-up for brand-new numbers
A fresh number blasting messages on day one is a dead number. Ramp new accounts gradually (0 → 500/day over a few days) so they build a normal-looking history first.
One-click secure public URL for the API (Cloudflare named tunnel)
Getting your self-hosted server reachable from n8n Cloud or a webhook shouldn't mean port-forwarding. Download the tunnel binary, click connect, get a stable hostname.
A stable, unique device identity per account
Each connected number should present one consistent device fingerprint for its whole life — assigned once, never shuffled. Rotating identities is exactly what gets numbers flagged.
Auto-cool a number when its risk spikes, resume when it settles
Instead of a blunt 24-hour lockout, back off automatically when the risk score climbs and lift the pause as it decays. Protect the number without babysitting it.
Quiet hours — never message people in the middle of the night
Set a nightly window per account where the server holds sends and drains the queue the next morning. Nothing feels more like a bot than a 3am "reminder".
High-Volume Mode for Pro (up to 5,000 sends a day)
For accounts that are properly warmed and behaving, let Pro users opt into much higher ceilings — up to 120/min and 5,000/day — behind a 72-hour warm-up and a clear acknowledgement.