Red Flag Scan

Single Text Message Analyzer

Paste one message or screenshot for a fast scan of coercion, manipulation, invalidation, pressure, and mixed signals before full-thread analysis.

What does a single-message red flag scan check for?

Red Flag Scan reads one pasted message or screenshot and flags coercion, manipulation, invalidation, pressure, and mixed signals. It is a fast first read rather than a relationship verdict: one message has no before and after, so the scan cannot show whether a pattern repeats or who changed first. A full history is what makes that visible.

  • Best input: a single message you want a second opinion on, pasted as text or as a screenshot.
  • What it cannot do: establish a pattern, a timeline, or a direction of change. That needs the whole thread.
  • Red Flag Scan is the fast path: paste one message or drop a screenshot and get a read on manipulation, pressure and invalidation patterns in that text.
  • It is deliberately narrower than Grey Mirror. A single message cannot show reciprocity, drift or repair, because those only exist across time.
  • Use it for triage on something that just landed. Use a full exported thread when the question is about the relationship rather than the message.
  • Findings describe language in the text supplied. They are not a diagnosis of the person who wrote it, and not evidence of intent.
  • If a message describes danger, that needs people rather than software: contact local emergency services or a domestic violence service.

Red Flag Scan is a free single text message analyzer. Paste one message, or upload a screenshot, and it returns a fast read on coercion, manipulation, invalidation, pressure and mixed signals. It is deliberately scoped to one message, which is both what makes it quick and what limits it: a single text has no before and after, so nothing it reports is a verdict about a relationship.

What one message can support is a read on the language itself -- whether a request is framed as a choice or as an obligation, whether disagreement is met with an answer or with an attack on the person, whether an apology takes responsibility or transfers it. What one message cannot support is a base rate. Whether this message is typical of how someone talks to you, or a departure from it, is only answerable across the whole history.

That is the honest handoff. If the question is about one text, the scan is the right tool and it is free. If the question is about a pattern -- is the effort mutual, did they cool off, when did it change, does conflict get repaired -- then the full-thread analyzer is the one built for it, because it reads the complete exported history and measures the behaviour across months rather than reading a moment.

Nothing pasted into the scan is used to train models. Screenshots are validated as real images before processing and are not retained beyond the request; the retention and deletion terms are published on the privacy page.

Start here

More from Grey Mirror

Related Guides

Keep the distinctions

A concerning sentence and a repeated pattern need different context

Visual guide updated

  1. Visible language

    A short exchange can show explicit threats, dismissal or a boundary being ignored.

  2. Repeated interaction

    A full history can show whether the concern recurs and how the other person responds to repair or limits.

  3. Real-world judgment

    Your safety and lived experience are not replaced by a score or an automated label.

Conceptual reading guide. This diagram explains what to inspect; it is not a customer report, diagnostic test or measured relationship score.
Use this guide on the question you came with

Use the amount of context your question needs. A tool should explain what it observed and what it cannot determine.