Stance

A claim whose typography IS its calibration. The agent states a claim and a confidence 0-1; the type sets itself from it -- weight from 300 (a hunch) to 900 (a conviction), size from 1.4rem to 3rem, tracking tightening and opacity rising with confidence -- under a thin confidence rail. Optional "because" (the reason) and "unless" (what would change the agent's mind, the line most opinion cards forget). A small slider lets the reader drag the confidence and watch the claim reset, which makes the mapping legible and honest. Same integer arithmetic on the server and in the slider script.

Surfaces
webgoogle-meet-stagegoogle-apps-script-webmcp-appspdfandroidslidespptx
Degraded on
SurfaceNote
google-chatRenders as a plain text card; typographic weighting and animation are lost
Live preview
stance
Claim
confidence 50%
Fields
FieldTypeRequired
claimstring (required). The claim, up to 200 characters.required
confidencenumber. 0-1 (values above 1 are read as percent). Default 0.5.optional
becausestring. The reason, one or two sentences.optional
unlessstring. What would change the agent''s mind; rendered as "I would change my mind if ...".optional
voice"display" | "serif" | "mono" (optional, default "display"). System font stacks.optional
theme"dark" | "light" (optional, default "dark")optional
accent"#rrggbb". Default "#38bdf8".optional
interactivebool. Show the try-it slider. Default true.optional
Example payload
{
  "type": "stance",
  "claim": "Claim"
}
ARD catalog entry
{
  "identifier": "urn:air:a2uicatalog.ai:atom:stance",
  "displayName": "Stance",
  "type": "application/vnd.a2ui.atom+json",
  "url": "https://a2uicatalog.ai/atoms/stance",
  "capabilities": [
    "web",
    "google-meet-stage",
    "google-apps-script-web",
    "mcp-apps",
    "pdf",
    "android",
    "slides",
    "pptx"
  ],
  "description": "A claim whose typography IS its calibration. The agent states a claim and a confidence 0-1; the type sets itself from it -- weight from 300 (a hunch) to 900 (a conviction), size from 1.4rem to 3rem, tracking tightening and opacity rising with confidence -- under a thin confidence rail. Optional \"because\" (the reason) and \"unless\" (what would change the agent's mind, the line most opinion cards forget). A small slider lets the reader drag the confidence and watch the claim reset, which makes the mapping legible and honest. Same integer arithmetic on the server and in the slider script.",
  "representativeQueries": [
    "show a claim set in type whose weight, size and tracking encode the stated confidence",
    "a claim whose typography is its calibration. the agent states a claim and a confidence 0-1; the type",
    "render a stance"
  ]
}
← Full ARD catalog
Deploy your own Google Apps Script renderer

The renderer is an open-source Google Apps Script web app. Deploy your own instance in 4 commands — you own the URL, no dependency on the demo endpoint.

git clone https://github.com/a2uicatalog/a2ui
cd a2ui/apps-script-surface/gas-wired-renderer
clasp push && clasp deploy
Full deploy guide →