Private page context

What private page context is

A page often knows why a visitor is opening chat: the plan they are comparing, the form they are completing, the product they selected, or the kind of interview the team wants the Stand-in to conduct. A prompt sends that context into the new Stand session without adding a visitor-authored message.

An AI Stand-in uses the prompt as session-specific guidance alongside team, site, and Stand-in instructions. A human rep can inspect it in the authenticated conversation view and History. The visitor does not see it in the public transcript.

Private means hidden from the visitor interface—not transient, encrypted for a special recipient, or exempt from conversation retention.

Supply context from a supported entry point

<stand-button
  hidden
  message="I can help compare plans. What would you like to know?"
  prompt="The visitor opened chat from the pricing comparison. Ask about team size and expected chat volume before recommending a plan."
>
  Ask about plans
</stand-button>

How Stand uses and stores it

• Stand attaches the prompt when it creates the session initiated by that entry point.
• For AI conversations, the prompt supplies facts and task guidance but does not become a visitor turn or determine the visitor’s language.
• For human conversations, authorized reps can read the context so they understand the page-specific intent.
• The context remains associated with the conversation according to the organization’s retention settings.
• The visitor’s actual form or textarea text should be sent separately as visitorMessage when it is meant to appear in the transcript.

Write context that changes the response

• State the page or workflow the visitor came from and any already-known selections.
• Tell the responder what to clarify before answering or recommending an option.
• Separate facts from instructions: for example, “Selected plan: Pro. Ask about team size before comparing.”
• Keep it specific to this activation. Reusable organization policy belongs in team, site, or Stand-in instructions.
• Do not claim facts the page cannot actually know or use prompt text as a substitute for authorization checks.

Visibility and privacy boundary

Timing and limitations

• Context applies to the session created by the activating button, card, or API request; it does not rewrite an already active conversation.
• It cannot authenticate the visitor or grant permission to reveal account-specific information.
• It does not replace knowledge sources for reusable product documentation or team instructions for organization-wide behavior.
• It does not appear as evidence that the visitor personally said or confirmed the supplied facts.

Related features

• Stand JavaScript API — The public window.StandChat API lets a website reveal its own chat entry points, inspect responder availability, open Stand with page-specific context, and identify a signed-in visitor.
• Stand button — The stand-button web component adds a compact, page-native chat invitation that appears only after Stand finds an Available rep or eligible Stand-in.
• Stand card — The stand-card web component turns authored page content into a contextual chat invitation with explicit activation, optional visitor input, responder identity, presentation themes, and spotlight behavior.