Your assistant sends a draft, such as a reply you are about to send, and gets back scores from 0 to 1 on three axes UpTrust calls Mirror: clear (direct and well structured), open (engages other views and admits uncertainty) and centered (grounded rather than reactive), with short notes on what drove each score. It does not check facts. The text goes to a third-party service provider to compute the scores.
Ask it like this
- “Does this reply sound defensive?”
- “Check the tone of this email before I send it.”
When another tool fits better
- Can I trust this source?
You want to know whether a source is reliable, not how a text reads.
Before you use it
- It needs the Read your trust signal permission, which UpTrust asks you to approve when you connect.
- It only reads. It changes nothing on UpTrust.
- In Claude, connect
https://mcp.uptrusting.com/mcp. It is not on the ChatGPT address,https://mcp.uptrusting.com/chatgpt. How to connect.
What your assistant is told
Your assistant decides when to use this tool from the title and description UpTrust’s server gives it. Here they are exactly as the server sends them.
- Tool name
mirror_score_text- Title
- How does this writing come across? (clarity and tone check)
- Description
Use this when the user wants to know how a piece of writing comes across before it is sent — a draft reply, your own output, a text they are reviewing — or to rewrite a draft they say feels "off", or to filter generated content that's technically correct but tonally wrong. It scores the text on three axes of tone and structure (UpTrust calls this check "Mirror"): `clear` (direct, well-structured), `open` (engages other perspectives, acknowledges uncertainty), and `centered` (emotionally grounded, non-reactive). It does not check facts or whether a source is reliable (that is is_trusted_by_principal). Each axis returns a [0, 1] score (null when the model abstained on that axis) and optionally 1-5 short markers explaining what drove the score. Data note: the supplied text is sent to a third-party service provider to compute the score.