Something already landed. You are holding it, reading it again, deciding what to do. That gap between what your gut wants to say and what you actually send is the moment reFrame™ was built for. Paste the message and see what is really in it.
Type your raw, honest thoughts below. Whether you're hurt, angry, frustrated, or confused — we'll help you express it in a way that builds connection instead of conflict. For best results, tell us who you're talking to and provide context about what they said.
Most tools only check what you are about to send. reFrame reads it either way. Flip the switch and watch the read change.
“It’s fine. Do whatever you want.”
You typed this. You have not hit send yet.
Passive aggression, with withdrawal underneath
They will not hear “it’s fine.” They will hear that the conversation is over, that they lost, and that it is going to cost them later. You have not ended the fight. You have postponed it and added interest.
“I’m too frustrated to land this well right now. I do care how it goes. Can we come back to it tonight?”
Educational, not a substitute for professional support. reFrame names patterns in text. It does not know your history, and it can be wrong about what someone meant.
reFrame is built for people who keep walking on eggshells in their own conversations. People who feel crazy after every text exchange. People who catch themselves second-guessing what they remember, or what they said, or what just happened. The patterns making you feel that way have names. reFrame names them. You are the only person you can't see, and this is the mirror for that blind spot.
reFrame detects the communication patterns that shape every conversation. The harmful ones research has linked to relationship damage: contempt, criticism, defensiveness, stonewalling, gaslighting, and others. And the healthy ones that build connection: repair attempts, validation, taking responsibility, and others. Ten harmful patterns, seven healthy ones. It names them in plain language and explains what each one is doing in the specific message you paste. The point is not to label the person who sent it or the person who is about to send it. The point is to make the pattern visible, so what's happening in the conversation stops being something that just happens to you, or through you.
Most communication tools only check what you are about to send. reFrame checks both directions. Paste a message someone sent you, and reFrame will tell you what pattern is in it. Paste a draft you are about to send, and reFrame will catch what your reader will receive.
The pause is the whole product. In the seconds between what your gut wants to say and what you actually send, there are three questions. Every reframed message is built on them:
The three questions are the R³ Framework. They came out of lived experience, not a lab, and they are the filter every reframed message passes through. The pattern detection underneath is separately informed by established relationship science. The questions work before you send a hard message, and they work for understanding a hard message you just received. Run them enough times and you stop needing us to ask them.
reFrame does not store your messages by default at any tier. No account is required to try it. Reading a message does mean sending it to the AI model that does the reading, and we would rather tell you that plainly than imply your words never go anywhere. What comes back to us, and all we keep, is the shape of it: which patterns were found, and how long the message was. Not the message.
This is a deliberate design choice. The category we work in includes domestic abuse survivors and people in custody disputes. Stored data in this category is a safety risk, not a feature. We treat it that way.
Guest reads need no signup at all. Paste something and see.
You have already read it ten times.
Read it once with the pattern showing.