Twilio SMS Delivery
Send transactional SMS with delivery tracking and opt-out handling that actually holds.
What it adds
An SMS delivery path with canonical phone numbers, per-message delivery state, and an enforced opt-out list.
What your agent is told to do
5
What your agent is told to do
5-
1
Identify the small set of messages that genuinely warrant SMS — verification codes, delivery alerts, appointment reminders — and route everything else to the app's existing email or in-app channels. SMS is expensive and intrusive.
-
2
Normalise every phone number to a canonical international form at the point it is captured, store it that way, and reject anything that cannot be parsed rather than attempting a send on a guess.
-
3
Send through the app's background-job system, and record a message row before dispatch that carries a stable identifier, the recipient, the purpose, and the current delivery state.
-
4
Check the opt-out list immediately before dispatch, not only when the message is queued. A message queued an hour ago must not go out to someone who has since replied to stop.
-
5
Do not treat a successful hand-off to the provider as delivery. Accepted means queued; the delivery outcome arrives later on a callback and the message state must reflect that distinction everywhere it is shown.
Edge cases it handles
8
Edge cases it handles
8- Numbers entered with local formatting, spaces, or a missing country code will fail or reach the wrong person. Normalise to a canonical international format on capture and store only that form.
- Standard opt-out keywords must unsubscribe the number immediately and permanently, and regional rules on quiet hours, sender identity, and consent differ by country. Encode the destination country's constraints rather than applying one policy everywhere.
- Queued, sent, delivered, undelivered, and failed are distinct states with different consequences. Model all of them, and never collapse undelivered into failed, because a temporarily unreachable handset is not a wrong number.
- Status callbacks arrive over a public endpoint. Verify their authenticity before trusting them, and make them idempotent, because the provider will resend the same event and events can arrive out of order.
- A job retried after a network timeout must not send the message again. Use a stable per-message key so a repeat attempt reconciles with the existing message instead of creating a second one.
- Long messages split into multiple segments and are billed per segment. Show the segment count where a message is composed, and be deliberate about characters that force a more expensive encoding.
- A per-recipient and per-account send ceiling must exist, so a loop or a retry storm cannot run up a bill or flood one person overnight.
- When the provider is degraded, the triggering action must still complete, the message must stay queued rather than being marked failed, and any user-facing copy must say the message is on its way rather than claiming it was delivered.
Definition of done
9
Definition of done
9- Every stored phone number is in a single canonical international format.
- Opt-out is checked at dispatch time and an opted-out number is never sent to again.
- Each message carries queued, sent, delivered, undelivered, or failed state, updated from verified callbacks.
- Status callbacks are authenticated and processing them repeatedly changes nothing.
- A retried job reconciles with the existing message rather than sending a duplicate.
- Per-recipient and per-account send ceilings are enforced.
- Provider downtime leaves messages queued and the triggering action successful.
- The feature matches the existing design system.
- No existing functionality is broken.
Related features
Firebase Cloud Messaging
Firebase Cloud Messaging
Deliver push through Firebase with registration tokens that stay valid as devices change.
What it does
A Firebase-backed push channel with server-side token lifecycle management, consistent payload handling across platforms, and safe handling of the routes a notification opens.
How it works
- 1 Store one record per device registration, tied to the account signed in at the time, the platform it came from, and when it was last confirmed working, and let one account hold many of them.
- 2 Send from the app's existing background-job system with the provider credentials held server-side, and treat every send as a job that can be retried rather than work done inline during a request.
- 3 Decide, per notification type, whether the payload is meant to be displayed by the platform or handled by the app, and keep that decision the same on every platform. The same event behaving as a visible alert on one platform and a silent update on another produces bugs nobody can reproduce.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/firebase-cloud-messaging
OneSignal Push Notifications
OneSignal Push Notifications
Send push notifications to the right devices, with user preferences respected.
What it does
A push delivery path through OneSignal covering subscription registration, audience targeting, per-channel preferences, and delivery outcomes recorded against the app's own records.
How it works
- 1 Find the events the app already notifies on by email or in-app, and treat push as another channel for those same events rather than a separate notification system with its own triggers.
- 2 Record the link between a push subscription and an account on the server, established at sign-in and severed at sign-out, so a device that changes hands does not keep receiving the previous user's notifications.
- 3 Hold the provider credentials server-side and send every message from the app's existing background-job system, so a slow or failing provider never blocks a web request.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/onesignal-push-notifications
Postmark Transactional Email
Postmark Transactional Email
Send transactional email through Postmark on the right stream, with bounces tracked.
What it does
A Postmark delivery path that separates message streams, honours suppression, and retains provider identifiers for diagnostics.
How it works
- 1 Put Postmark behind the app's existing mail layer as one adapter, so no calling code constructs a provider-specific payload. This entry is an alternative to the other transactional email integrations, not an addition — pick one adapter and keep the mail layer identical.
- 2 Classify every outbound message as transactional or broadcast and bind that classification to the matching message stream at the mail layer, so the choice is made once and cannot be got wrong at an individual send site.
- 3 Send from the app's background-job system and persist the provider's message identifier on the record the moment it comes back, so a support question about one email can be answered without guessing.
Copy the prompt
No account needed
Add this feature to my app:
https://addthisfeature.com/x/postmark-transactional-email
How it works
-
1
Copy the link
Grab the Markdown instruction URL for this feature.
-
2
Give it to your AI
Paste it into Claude Code, Cursor, v0, Lovable — whatever you build with.
-
3
It inspects, then implements
Your agent reads your existing app first, then adds the feature to fit it.
Works with your stack
These instructions are written to adapt. They tell the agent to detect your framework, match your existing design system, and reuse what you already have — rather than assuming a particular stack.
Need it tighter than that? Customize the feature and tell it exactly what you're running.