Check Your Scheduled WhatsApp Messages Went Out, with Claude (2026)

You scheduled the reminders on Sunday night. It is Wednesday, and you cannot remember which ones are still waiting, which went out, and whether the one to the supplier ever got read.

You can open each chat and look. Or you can ask Claude. With the Blueticks WhatsApp MCP connector, Claude can list your scheduled messages, look at one in detail, and look up delivery and read receipts. The prompt below is written to only read, but the same tool can also send, edit and cancel, so you approve each call.

What can Claude see about a scheduled message?

Claude uses the scheduled_messages tool. Three of its actions do the work in a check:

The list covers your whole account, so a reminder you set up in the Blueticks app shows up next to one Claude scheduled for you.

The main statuses are pending (accepted, not sent yet), confirmed, received, read, played and failed. Other values can appear on older or stuck rows, so ask Claude to report any status it does not recognise. A row also carries the planned send time (sendAt), the actual time once it went out (confirmedAt), and a failure reason when there is one.

You never type the action names. You describe what you want in plain words. The names help when an answer looks wrong and you want to say "list it again with sent messages included."

What do you need before you start?

A linked WhatsApp number in Blueticks and the connector added to Claude. If you have not done that, how to connect WhatsApp to Claude covers it. The connector URL is https://api.blueticks.co/mcp and you sign in with OAuth.

This check uses scheduled_messages, plus utils for today's date. The full list is in the WhatsApp MCP tools guide.

The prompt: what is waiting and what went out

Paste this into a normal Claude chat:

Using Blueticks, check my scheduled WhatsApp messages. Get today's date first. First list everything still pending, following the paging to the last page, with the recipient name, the planned send time and the status. Then list the messages that were sent, including sent messages, and follow the paging to the last page. Keep only rows whose status is not pending and whose confirmedAt is in the last 7 days, and tell me for each whether it was received or read. Then list anything with the status failed, including sent messages, following the paging, with the failure reason. If a list returns a nextCursor or hasMore: true that you could not follow to the end, tell me the answer is incomplete. Do not quote the message text. Do not send, edit, reschedule or cancel anything.

Behind it, Claude makes several calls, more if you have a long history:

1. Date. utils returns the current date, so "7 days back" is a real timestamp.
2. Pending. scheduled_messages runs list with status set to pending.
3. Sent. A second list with sent messages included. The list has no date filter and is ordered by when each message was created, not when it was sent. So Claude follows nextCursor to the last page and applies the 7-day cut itself, on the time the message actually went out.
4. Receipts. ack takes the message keys from those rows (the row's waMessageKey, in its string form _serialized) and returns the receipts.
5. Failed. A third list with status set to failed.

A sample answer, with invented names:

Still pending (2)

• Dana, tomorrow 09:00, pending
• Supplier group, Friday 08:30, pending

Sent in the last 7 days (3)

• Lior, Sun 20:00: read
• Landlord, Mon 09:00: received, not read
• Team group, Tue 07:45: received

Failed (1)

• Noa, Mon 18:30: reason given

The prompt asks Claude not to quote the text for a reason: the list rows carry the message body, so Claude can read it. Naming recipient, time and status is enough for a check, and it keeps private wording out of the answer you might paste elsewhere.

Why do sent messages seem to disappear from the list?

They are hidden by default: a plain list leaves out messages that were already sent. To see what went out, the call needs include_sent set to true. That is why the prompt says "including sent messages," so a list that looks empty after a busy week does not mean nothing happened.

Know these two things:

• Cancelled messages never come back. A message you cancelled is not listed, even with sent messages included.
• A sent message has a key. The WhatsApp message key is empty until the engine dispatches the message, and with include_sent, a sent message is listed only once it has one.

What do the receipts mean?

ack returns one number per message key:

Two results look worse than they are.

• found: false. The key was not found in the engine's store. That does not mean it was not delivered. Check the status in the list or get row instead.
• null with found: true. The engine could not determine a receipt. Again, not proof of failure.

A message can be received and never show as read. That is ordinary WhatsApp behaviour, so "received, not read" is not a sign of a problem.

Other checks that work the same way

Swap one line of the prompt for a different check:

• One recipient. "Everything I scheduled for Dana, pending and sent." This uses the chat_id filter.
• Only the problems. "List failed messages and the reason for each." This uses status set to failed.
• One message you just scheduled. Ask Claude to look it up by its id until it has a key, then check the receipt.
• Before a trip. "What is pending for the next 7 days?" Then fix the odd one by asking Claude to move or cancel it, after you confirm the exact message.

What can go wrong?

• The engine is offline. The pending and failed lists still work. Only the receipt lookup reads from your connected engine, and it can fail while the engine is down. Check the engine status in Blueticks, reconnect WhatsApp if it is offline, and run the check again.
• A long list. The list pages its results and has no date filter, so a big history means more pages. Every recipient of a campaign counts as a message, so a big campaign makes the sent list long. Ask for one recipient, or for only the pending and failed messages.
• A similar name. If two contacts share a first name, give Claude the exact name or the number.

Is it safe to let Claude look at your scheduled messages?

The check reads. But the same scheduled_messages tool also sends, edits and cancels, so you decide what runs. Anthropic's custom connector guide says that during Research, Claude can use connector tools without further approval, so run this check in a normal chat instead. When Claude asks for approval, read the action name before you allow it, and avoid "Allow always" for a connector that can send.

The line "do not send, edit, reschedule or cancel anything" is only a request to Claude. In a normal chat, your approval of each call is what stops an action.

Related guides

To schedule the messages in the first place, see scheduling WhatsApp messages from Claude. If a message did not go out and you want to know why, read why scheduled WhatsApp messages are not sending. Plans and prices are on the pricing page, and you can start free.

Frequently asked questions

Can Claude tell me if a scheduled WhatsApp message was sent? Yes. With the Blueticks connector, Claude can list your messages by status and look up delivery and read receipts for the ones that went out.

Why is my sent message missing from the list? Sent messages are left out unless the list is asked to include them. Cancelled messages are never listed.

Can it tell me if someone read my message? Where a read receipt exists, yes. A message can be received without ever showing as read, so a missing read receipt does not mean it failed.

Does the check send or change anything? Not if you run the prompt as written: it uses read actions. In a normal chat, approve every call and skip "Allow always."

What if the receipt says not found? Check the message status in the list instead. A key that is not found is not the same as a message that was not delivered.