See exactly what ChatMend reveals
This is a complete, anonymized analysis of a fictional conversation. Real reports follow this same structure, built entirely from the messages you provide.
Fictional example. No real person or conversation is shown here. When you run your own analysis, the conversation text is processed to generate your report and isn't retained afterward — saved results live in your account, and you can delete them anytime.
The whole report, in one card
Before the detail, ChatMend distills the dynamic into a single snapshot: the dominant pattern, how confident the read is, the evidence behind it, and whether repair is actually happening.
- One headline pattern, not a wall of scores
- A confidence level you can trust or question
- Evidence pulled straight from the messages
Relationship Health
This conversation shows strong connection and genuine emotional interest, alongside repeated ambiguity after vulnerable moments. The core issue is not a lack of care — it is inconsistency in the moments right after emotional closeness.
How each person tends to connect
Presented as signals, not diagnoses. These are tendencies the conversation suggests — not fixed labels.
Seeks reassurance after silences and escalates emotional bids when responses slow down.
Creates distance after emotionally intense or vulnerable exchanges, then re-engages warmly later.
A pursue-withdraw cycle around closeness
Plotted across anxiety and avoidance, the two people sit in different corners. Alex leans anxious — moving toward connection under stress. Jordan leans avoidant — moving away to manage overwhelm. The dashed line is the loop that forms between them.
Vulnerable message
AlexDelay / withdrawal
JordanReassurance seeking
AlexConflict
BothRepair attempt
JordanA vulnerable or emotionally warm message is shared.
Jordan goes quiet for several hours without explaining why.
Alex follows up for reassurance, then repair happens the next day.
The cycle is not driven by lack of interest. It is a timing mismatch: one person moves toward connection under stress, the other moves away. Naming the rhythm is usually more useful than assigning fault.
Every finding links to the messages
No vague interpretations. Each signal points to what actually happened in the thread.
“Consistent 3-5 hour response delays following vulnerability-based messages, paired with warm re-engagement the next day. This pattern appeared 12 times across the conversation.”
“After a delayed reply, Alex sent 2-4 follow-up messages within an hour (“did I say something?” / “it’s fine, ignore me”) in 9 of 12 instances.”
“In 8 of 12 cycles the connection returned to warmth within 24 hours, usually initiated by Alex and accepted by Jordan. Repair happens — it is just slow and one-sided.”
Reconnection after distance is initiated almost entirely by one person. Over time this can build quiet resentment and burnout.
Distance is rarely explained, so the other person fills the silence with worst-case interpretations.
Rapid follow-up messages during anxiety can increase the very distance they are trying to close.
The pattern formed gradually
High message frequency, fast replies on both sides, lots of playful and affectionate exchanges. Secure-looking baseline.
First vulnerable conversations appear. The first response delays follow, and the pursue-withdraw rhythm begins to form.
The cycle is now recognizable and repeating. Warmth remains, but the timing of distance is predictable around emotional closeness.
Grounded steps, not generic advice
Suggestions are tailored to the actual pattern in this conversation.
Name the pattern without blame
Describe the timing, not the character. The goal is a shared observation, not an accusation.
Make repair a two-way habit
If one person always initiates reconnection, agree on a small signal the other can use to reach back first.
Reduce the reassurance loop
When a reply is slow, one clear message lands better than several. Distance often eases when the bid is calm.
“I’ve noticed that when I share something personal, you sometimes go quiet for a while, and I start to worry. I’d like to understand what happens for you in those moments — not to fix it, just to get it.”
This opens dialogue without blame — an invitation to understand, not a verdict.
ChatMend organizes patterns and language to support reflection and therapy preparation. It does not diagnose and is not a replacement for a licensed professional. AI analysis can be wrong and cannot see the full context of your relationship.
Run this on your own conversation
Paste a thread or upload a file. You will get a report like this one — built only from your messages.