Overview
Auto-fill
Typeahead suggestions that auto-complete user prompts
autocompletetypeaheadsuggestions
Docs
Auto-fill accelerates prompt composition by suggesting completions as the user types. Drawing from recent queries, popular prompts, or AI-powered predictions, it reduces the keystrokes needed to express intent and helps users discover effective prompt patterns they might not have thought of. The key challenge is balancing helpfulness with unobtrusiveness — suggestions should appear quickly and disappear cleanly without interrupting the user's flow.
Variants
- Dropdown suggestions — A list of completion options that appears below the input field, filtered in real time as the user types.
- Inline ghost text — Semi-transparent text that extends the cursor position, accepted with Tab or dismissed by continuing to type.
- Categorized suggestions — Completions grouped by source (recent, popular, AI-predicted) with labels indicating provenance.
Use Cases
- Search interfaces where users benefit from query suggestions
- Chat interfaces that accelerate common prompt patterns
- Code editors providing AI-powered line completions
- Form fields that auto-populate based on context or history
- Command palettes that suggest actions matching partial input
Best Practices
- Show suggestions after a short debounce (150-300ms) to avoid flickering during fast typing.
- Highlight the matching portion of each suggestion so users can see why it was offered.
- Support keyboard navigation (arrow keys, Enter to select, Escape to dismiss).
- Limit visible suggestions to 4-6 to keep the dropdown scannable.
- Indicate the source of each suggestion (history, popular, AI-generated) for transparency.