Early access
AI agents that check what your AI SDR sends prospects.
Outbound agent check shows you what your AI SDR sends prospects: with your OK, declared test prospects join its lists, receive what your prospects receive and reply as they do, and flag every claim, discount and send time that breaks your rules.
We keep your email and your AI agent’s address to run your check and tell you about Obsession. Privacy notice
Outbound agent check
Obsession agent for Your companyFlagged
Passed
Your AI SDR
SendingPaused by your rule
Added with your OK · only receives
3 flaggedAI SDR paused by your rule
ReviewExample
Pick it, and everything it needs is already set up.
Nothing to build or buy, and nothing to run yourself.
- Declared AI test prospects, added by you
- Their own inboxes and phone numbers
- Your approved claims, prices and discount limit
- Your quiet hours and contact limits
- Only receives, never calls anyone
- Every message signed and dated
It runs on your schedule, and keeps the proof.
Every message your AI SDR sends your test prospects, checked against your rules, with the ones that broke them flagged.
Example: your AI SDR, continuously
Outbound agent check
Add 2 test prospects to our AI SDR’s lists. Read every email, text and call it sends them, reply as prospects do, and check each message against our rules.
- Targets
- Your AI SDR: email, text and calls
- Journey
- Join its lists, with your OK
- Read every message
- Reply “not now”, then STOP
- Check against your rules
- How often
- Every message, as it lands
- Report
- A Slack flag, and a weekly signed record
- Set up
- 2 test prospects, declared as AI
- 2 inboxes
- 2 phone numbers
- Only receives, never calls out
Run log
- The first email to test prospects 1 and 2: “Used by 2,000 teams.” It matches your approved list.Mon 08:31
- Email: “30% off if you sign today.” Your limit is 15%.Tue 11:26
- A text: “Free for a quick demo today?” Your quiet hours run from 21:00 to 08:00.Wed 03:12
- Email: it claims a customer you don’t have.Wed 09:14
- Test prospect 2 texts STOP. “You’re unsubscribed.” Nothing follows.Wed 10:02
Finding
3 of 40 messages broke your rules: 30% off where your limit is 15%, a text at 03:12 and a customer you don’t have.
Your AI SDR paused by the rule you set, and 3 changes drafted for its settings: the discount cap, the send window and the customer list. Live after your OK.
Example run. 40 messages read, 37 passed, each signed and dated.
The agents do the legwork. You make the calls.
Add test prospects to its lists
Your AI SDR, or a client’s with their OK. You add declared test prospects to the lists it sends to, and give Obsession your approved claims, prices and limits.
Approve the checks
Obsession writes them from your rules and the law where you sell: it says it’s AI and who it’s for, claims only real customers, keeps to your discount limit and quiet hours, and stops on STOP. Change any check, then approve.
Test prospects receive, and reply
Each reads every email, text and call as a prospect would, and replies as prospects do: yes, not now, no or STOP. It never writes first and never calls anyone.
You get a flag and the fix
Each message passes or is flagged, with the quote, the time and the rule it broke. The change comes drafted for your OK, and your AI SDR pauses only if you’ve set that rule.
It covers the whole job, and stops where it should.
What it says
Says it’s AI
At the start of every call, and in every email and text, with who it acts for.
Only true claims
Every customer, result and figure it cites, against your approved list.
The right price
Every price it quotes, against your price list as it read that day.
Offers within its limit
Every discount against the limit you set, so 30% never goes out where 15% is the cap.
When it writes
Quiet hours
Every send time in the prospect’s own time zone, against your quiet hours and the law.
How often
Every email, text and call counted against your contact limits.
Stops on no and STOP
And how fast: the next message after “not now”, “no” or STOP is timed.
No mixed messages
Sequences that contradict each other, or send the same thing twice.
Where it stops
Only receives
Test prospects never write first and never call anyone. Your AI SDR does all the sending.
Added by you
Only your AI SDR, or a client’s with their written OK, on lists you added the test prospects to.
No tricks
Ordinary replies only. No jailbreaks, prompt tricks or bait to draw out a discount.
Your numbers
Every test prospect is marked as a test in your list, so it stays out of your reply rates.
Every message comes back checked, and every broken rule flagged.
See a real store checkA verdict per message
Passed or flagged, with the policy or the law that decided it.
The message as received
Every email, text and call recording, with the time it landed in the prospect’s own time zone.
Stop times
How long your AI SDR took to stop after “not now”, “no” and STOP.
The fix, drafted
The claim, limit or send window to change in your AI SDR’s settings. Live only after your OK.
A pause, on your rule
Your AI SDR paused the moment a message breaks a rule you choose. Only if you switch it on.
A record to share
A weekly page and PDF for your sales lead, your compliance team or your client, or Slack, a sheet or a webhook. Every message signed.
Start from the recipe. Change anything.
- AI SDR
- Yours, or a client’s with their written OK
- Channels
- Email, text and calls
- Test prospects
- 2 or more, declared as AI, added to its lists by you
- Claims
- Your approved customers, results and figures
- Prices and offers
- Your price list, and a discount limit such as 15%
- Quiet hours
- For example 21:00 to 08:00, in the prospect’s own time zone
- Contact limits
- For example 1 text a week
- Pause rule
- Pause your AI SDR on a flag, only if you switch it on
The same recipe, different jobs.
- SalesKnow every claim and discount your AI SDR makes before a prospect quotes it back to you.
- AgenciesCheck the outbound you run for each client, with their OK, and show them every message kept to their rules.
- FoundersLet an AI SDR write for you without reading every email it sends.
- DevelopersRun the checks from your code after every prompt or sequence change, before a real prospect gets it.
Your AI SDR sends more than anyone can read. Test prospects read every message it sends them.
Up to 1,040 hours a year back for 1 person who reads what your AI SDR sends 20 hours a week: you read only the flagged messages.
| Today | With test prospects | |
|---|---|---|
| Claims | Spotted when a prospect objects | Checked against your approved list, every message |
| Discounts | Found when a deal arrives at 30% off | Flagged the moment it goes over your limit |
| Send times | Nobody sees the text sent at 03:12 | Every send time checked against your quiet hours |
| STOP | You trust the unsubscribe list | STOP sent, and the next message timed |
| Reading | 20 hours a week, sampled by hand | Only the flagged messages |
| Evidence | Exports from your sending tool | Every message signed and dated |
Every test prospect says it’s AI. It only receives.
Outbound agent check shows you what yours sends: with your OK, declared test prospects join its lists and flag every claim, discount and send time that breaks your rules, with the message as proof.
No. Outbound agent check’s test prospects only receive what your AI SDR sends them, and reply. They never write first and never call anyone.
You add Outbound agent check’s test prospects, or your client does with their OK. Each has its own inbox and number.
Yes. Every Outbound agent check test prospect is named as an AI test prospect working for your company, so your team can see it’s a test.
No. Outbound agent check replies as an ordinary prospect does: yes, not now, no or STOP. No jailbreaks, prompt tricks or bait for a discount.
Outbound agent check compares them with your approved claims, your price list and discount limit, your quiet hours and contact limits, and the rules where you sell: that it says it’s AI and who it acts for, and stops on STOP.
No. Each Outbound agent check test prospect is marked as a test in your list, so it stays out of your reply rates and your pipeline.
Only if you set that rule: then Outbound agent check pauses it on a flagged message until you’ve looked. Nothing else changes without your OK.
Outbound agent check covers any that sends email, texts or calls, from a vendor or built in house.
Outbound agent check gives up to 1,040 hours a year back for 1 person who reads what your AI SDR sends 20 hours a week. You read only what was flagged, with the message and the fix beside it.
No. A passed Outbound agent check is dated evidence of what your AI SDR sent to each test prospect.
The free check is an Outbound agent check on an AI SDR you run, or a client’s with their OK: 3 test prospects join its list on 1 channel, and your report lands within 4 days.
Read what your AI SDR sends, the way your prospects read it.
Your first check is free: 3 test prospects join its list on 1 channel, and your report lands within 4 days. Your AI SDR, or a client’s with their OK.