Web Search Results
The query, then each page found with its own state: queued, reading, read or failed. Past a few rows the rest go behind "+7 more", and when the search is done it folds to "Read 7 pages from 5 sites" with the site logos.
The decision inside it
A failed page stays in the list
The reader should know the agent tried the paywalled source and could not get in. Dropping it would make the answer look better sourced than it is.
When to use it
Use Web Search Results while an agent searches the web and reads pages: each page shows queued, reading, read or failed, and the list folds to one line when the search is done. Update results and status as your agent streams events.
Installation
How you get it
Paste this into Claude Code, Cursor, Codex or any coding agent. It installs the file and uses it as it is, instead of writing a new one.
Use Layro's Web Search Results for this.
Install it: npx shadcn@latest add @layro/web-search-results
(the @layro registry is set up in components.json, with LAYRO_KEY in .env.local)
Import it from @/components/ui/web-search-results.
Do not rewrite or restyle it. Pass my data through its props
and keep its built-in states.For the full brief with every prop and the mistakes assistants usually make, use Copy prompt above. Pro components install through the @layro registry with your CLI token: see the Namespaced tab.
Usage
import { WebSearchResults, type SearchResult } from "@/components/ui/web-search-results";
export function AgentSearch({ query, results, done }: { query: string; results: SearchResult[]; done: boolean }) {
return <WebSearchResults query={query} results={results} status={done ? "done" : "reading"} maxVisible={5} />;
}API reference
“Copy prompt” above puts a full brief on your clipboard — install command, dependencies, props, and the mistakes an assistant usually makes with this component.
Props
Read from the component’s own TypeScript interface when the site is built, so this table cannot drift from the file you install.
| Prop | Type | Notes |
|---|---|---|
queryrequired | string | What was searched for, as the agent typed it. |
resultsrequired | SearchResult[] | The pages found so far. |
status | "searching" | "reading" | "done" | "error" | The search as a whole. Default "reading". |
maxVisible | number | Rows shown before "+N more". Default 5. |
variant | "default" | "compact" | "compact" drops snippets and tightens rows, for a narrow chat column. Default "default". |
onResultClick | (result: SearchResult) => void | Called when a row is pressed. Without it, rows are links that open in a new tab. |
defaultExpanded | boolean | Start with the list open even when done. Default false: a finished search folds to one line. |
look | Look | Layered rests the list in a tray with the totals in the strip below. Flat is one card. Default layered. |
Dependencies
npm packages
clsxlucide-reacttailwind-mergeLayro foundations, installed alongside it
layro-themeStates
All of them are drawn by the component. None is left for you to design.
- Searching
- Reading
- Done
- Failed page
- Compact
Accessibility
- Keyboard reachable throughout, in visual order
- Focus trapped in overlays, Escape closes them
- Focus returns to whatever opened it
- Live regions announce what changed, not that something did
- Honours prefers-reduced-motion
Responsive behavior
- Sizes to its container, not the viewport
- Usable from 320px up
- Pointer-based interaction, so trackpad and touch both work
- No hover-only affordances
Used in a screenAll screens
Lifetime access
Every component, and everything we ship next.
- All 232 components, source included
- Every template, and everything released after you buy
- shadcn CLI or download — React 18/19, Tailwind v4
- One payment. No subscription.
More in AI AgentAll AI Agent components
Agent Cursors
AI agents working in a document like multiplayer teammates, each with a named cursor.
Agent Workspace
A whole agent screen: task in, plan, tool calls, approval and a summary out.
Agent Run Status
The top of a run: state, current step, a live clock, and Pause and Stop.
Task Plan
The agent's plan as steps that tick off live, with the risky one flagged ahead.
Tool Call
One tool the agent used: what it sent, what came back, how long it took.
Approval Request
The agent stops and asks first, showing the exact thing it wants to do.