Relationship clarity, backed by evidence

The truth was in the texts the whole time.

Upload a real conversation and get an honest, evidence-based read on the red flags, communication patterns, and attachment styles hiding in how they talk to you. Instant. Private. Built on AI that never stores a single word.

πŸ”’ No sign-up to start Β· Nothing stored Β· Delete anytime
A
Alex
iMessage
Today 9:14 PM
I never said I'd go. You always make stuff up.
🚩 Gaslighting · High
You texted me yesterday that you wanted to though
Wow, so now I'm a liar? This is why nobody can deal with you.
🚩 Blame-shifting · High
I'm not calling you a liar
Good. Now drop it. I don't want to hear about it again.
🚩 Stonewalling · High
82
Several red flagsRed-flag score
PRIVATE BY DESIGNβ€’NO-RETENTION AIβ€’EVIDENCE-BASEDβ€’NOTHING STOREDβ€’DELETE ANYTIME
Sound familiar?

You're not crazy. You're just missing the pattern.

When you're inside it, the same messages can mean ten different things. From the outside, the pattern is obvious.

πŸŒ€

One minute they adore you, the next you're the problem. You can never tell which version you'll get.

πŸŒ™

You re-read the same texts at 2am, trying to figure out what's actually real.

🧭

Everyone says "trust your gut" β€” but your gut keeps getting talked out of it.

The patterns are sitting right there in your messages. We help you see them clearly β€” with the receipts.

How it works

Clarity in three steps. About 60 seconds.

1

Export your chat

In WhatsApp, open the chat β†’ Export Chat β†’ Without Media. You get a simple .txt file. Takes ten seconds.

2

Upload it privately

Drop the file in. It's analyzed in-memory on a no-retention AI and never written to a database. Ever.

3

Get your read

A scored report with the exact quotes behind every flag β€” plus a chat that has read the entire conversation.

What you get

Not vague vibes. A real, evidence-based read.

🚩

Red-flag score

A clear 0–100 read on how concerning the patterns are β€” not a guess, a measurement.

🧾

Evidence, not opinions

Every single flag is backed by the actual messages. You see exactly why.

🧠

Attachment styles

See whether they β€” and you β€” lean secure, anxious, or avoidant.

βš–οΈ

Honest about both sides

It reads your patterns too. Built for growth, not ammunition.

πŸ’¬

Chat with your texts

Ask "is this gaslighting?" and get a grounded answer, with quotes.

πŸ”’

Truly private

No-retention AI. Your messages are never stored or used to train anything.

See it in action

This is what clarity looks like.

A real report names the pattern, scores the severity, and shows you the messages behind it.

A
Alex
iMessage
Yesterday 11:02 PM
Can we talk about what happened last night?
You're too sensitive. I was the best thing that ever happened to you.
🚩 Contempt · High
That really hurt though.
Fine. I'm sorry you feel that way. Happy now?
🚩 Non-apology · Moderate
πŸ’¬ Ask anything
"Is this gaslighting, or am I overreacting?"
You're not overreacting. Dismissing your feelings as "too sensitive," then offering a non-apology, is a recognizable pattern, and it shows up more than once in this conversation.
Privacy isn't a feature β€” it's the foundation

Your most private conversations stay private.

This is the most personal data you own. We treat it that way.

  • No-retention AI. Inference runs on a model contractually bound to never store or train on your messages.
  • Processed, then gone. Your chat is analyzed in-memory for the request and never written to a database.
  • No account to start. You don't hand over an identity to get your read.
  • You're in control. Nothing is kept, so there's nothing to leak β€” and you can walk away anytime.
πŸ”’
Never stored.
Never trained on.
Your words stay yours.
Pricing

Start free. Go deeper for the price of a coffee.

Free
$0

See if there's something here.

  • βœ“ Your red-flag score
  • βœ“ Your first flag, with evidence
  • βœ“ Fully private
Get my score
Most popular
Pro
$3.99/mo

The whole picture.

  • βœ“ Every red flag, fully detailed
  • βœ“ Both attachment-style reads
  • βœ“ Your own patterns, honestly
  • βœ“ Unlimited chat-with-your-texts
  • βœ“ Re-run anytime as things change
Start free, upgrade anytime
Your free check

Ready to stop second-guessing?

Drop in a WhatsApp export and get your read in about a minute. Private the whole way.

How do I export a WhatsApp chat?
  1. Open the chat in WhatsApp
  2. Tap the contact name β†’ Export Chat
  3. Choose Without Media
  4. Save / AirDrop the .txt file, then upload it here

πŸ”’ Analysis runs on a no-retention AI. Nothing you upload is stored or used to train anything, ever.

Questions

Everything you're probably wondering.

Is this actually private?

Yes. Your conversation is analyzed in-memory on a no-retention AI and is never written to a database or used for training. There's no account required to start, and because nothing is stored, there's nothing to leak.

Is this a medical or psychological diagnosis?

No. It's an informational read on communication patterns, backed by the actual messages β€” not therapy and not a clinical diagnosis. It describes behaviors and patterns; what you do with that clarity is up to you.

What if I only use iMessage?

WhatsApp export works today. iMessage and screenshot uploads are on the way β€” the analysis engine already supports adding new sources.

Will the other person know?

No. You're analyzing your own conversation on your own device. Nobody is contacted or notified.

How accurate is it?

Every flag is tied to specific quotes, so you can judge each one yourself. It surfaces patterns with evidence β€” it doesn't pretend to read minds. You stay the decision-maker.

Can I delete my data?

There's nothing to delete β€” your messages are never stored. Close the tab and it's gone.

Trust your gut. Then check the receipts.

Get an honest, evidence-based read in about a minute β€” completely private.

Check my texts β€” free

Which one is you?

We found these people in the chat. Tap your name so we know whose side is whose.

🚩

Reading between the lines…

Scanning for patterns, attachment styles, and red flags.

Hmm, that didn't work