Pattern library

Adds a second verification step (TOTP code, SMS, or push notification) after password to dramatically reduce account takeover risk.

Standard email-and-password sign-in. The user enters credentials, the system verifies them, and a session is created on success.

Passwordless sign-in. User enters email, gets a one-time link, clicks it, and is signed in. No password to remember.

Let users sign in using their existing Google account. No new password to remember; faster signup.

Lets users recover access when they forget their password, via a time-limited email link to set a new one.

A turn-by-turn conversation with an AI, where each user message and AI response is added to a growing history that the AI sees on the next turn.

Detects when AI cannot help (low confidence, sensitive topic, explicit user request) and seamlessly transfers the conversation to a human agent with full context.

Adds your private documents to the AI's knowledge by finding relevant chunks via embedding similarity and injecting them into the prompt before the AI answers.

Shows AI response token-by-token as it generates instead of waiting for the full reply, dramatically improving perceived speed.

Lets the AI invoke functions you define (lookup data, send email, create record) instead of just generating text. Foundation of "agentic" behavior.

Lets users add products, change quantities, and review their selection before checkout. Persists across sessions.

Takes payment from the user via a payment provider (Stripe, LINE Pay, 토스), confirms the order, and triggers fulfillment.

Lets users apply a discount code at checkout (percentage off, fixed amount, free shipping) with rules around usage limits and expiry.

Lets the user request a refund (full or partial), evaluates the request against your policy, processes via the payment provider, and updates the order status.

Recurring payments on a schedule (monthly, yearly), with self-serve plan changes, cancellations, and dunning when payments fail.

A chronological stream showing what happened on a resource or for a user — who changed what, when. Foundation of audit trails and "what's new" sections.

A submitter sends a resource (document, expense, design) to one or more approvers, who can approve, reject with reason, or request changes.

Lets users add text comments to a resource (post, document, image), see others' comments, and edit/delete their own.

Lets a user reference another user by typing @ followed by a name, with autocomplete. Mentioned users get notified.

Multiple users see each other's presence and changes to a shared resource in real-time (cursor positions, edits, selections), like Figma or Google Docs.