Dashboard Overview
The screen an app opens on: shell, header, KPI row, chart and activity feed — numbers first, shape second, changes last.
Five ways to reach a model without the cursor leaving the sentence — slash menu, inline completion, selection toolbar, action menu, property inspector — and the two surfaces that say where the answer came from.
app/ai-editor-page.tsxnpx hoverlab add ai-editor-page
Or over MCP, from your editor's agent — no account needed.
Free to read, copy and install, for personal and non-commercial projects. Shipping it in client work or a paid product needs Pro ($79 once). The source lands in your repo and stops being ours — no attribution, nothing to upgrade.
Five ways to reach the model without leaving the sentence you were writing — and the two surfaces that say where its answer came from.
Type / for actions
Escape removes the slash and anything typed after it — it does not leave you to clean up.
Keep typing — completions appear after a short pause.
The renewal conversation should start ninety days out, not thirty. By the time a customer is thirty days from renewal they have already formed a view, and the only lever left is discount. Teams that open the conversation at ninety days spend it on scope instead of price, and close at a materially higher rate — the difference in our own book was eleven points last year.
Select any text above to raise the toolbar. Shift-Arrow works too.
Every row says where the result lands. An AI action that silently replaces the paragraph someone spent ten minutes on is the fastest way to have the feature turned off.
Burnt honey & sea salt
Slow-churned with wildflower honey, finished with flaked salt.
Grounding
The interesting half of a retrieval system is not what it does when it finds the answer. It is what it does when it does not.
Oversight
"A human is in the loop" is a claim with no shape. These are the four points where one really is, and each is a thing the system will not do without them.
The real page, rendered in your current theme — every section below is a live block, not a screenshot.
Only want one section? Open it and copy that block instead — browse Inline AI Actions.
/**
* AI inside a document, rather than beside one.
*
* The chat window is the easy shape and the wrong one for editing. Every
* surface on this page shares a single constraint that a chat panel does
* not have: the user is in the middle of writing something, and the model
* has to arrive without taking the cursor, the selection or the scroll
* position away from them.
*
* That constraint is why the order runs from least to most interruptive.
*
* slash menu invoked by typing, so focus never leaves the editor
* inline a completion in the text itself, dismissed by typing
* selection a toolbar that only exists while a selection does
* action menu an explicit, deliberate list of transformations
* inspector the model edits real properties, each one attributed
* and reversible
*
* <GroundingSplit> and <HumanOversightSplit> close the page and they are a
* different kind of thing: not editing surfaces but the two answers to
* "where did that come from" and "who is accountable for it". They belong
* on this page rather than on the retrieval console because the person who
* needs them is the one about to publish the paragraph, not the one
* debugging the index.
*
* The inspector is the most interruptive and sits last among the editing
* surfaces for a reason worth stating: per-property provenance is what
* makes a bulk AI edit reversible, and reversibility is the thing that lets
* someone accept a suggestion without reading all of it first.
*/
import * as React from 'react'
import { AiSlashMenu } from '@/components/ai-slash-menu'
import { AiInlineSuggestion } from '@/components/ai-inline-suggestion'
import { SelectionAiToolbar } from '@/components/selection-ai-toolbar'
import { AiActionMenu } from '@/components/ai-action-menu'
import { AiInspectorPanel } from '@/components/ai-inspector-panel'
import { GroundingSplit } from '@/components/grounding-split'
import { HumanOversightSplit } from '@/components/human-oversight-split'
export default function AiEditorPage() {
return (
<main className="min-h-screen bg-background text-foreground">
<section className="mx-auto w-full max-w-5xl px-6 pb-2 pt-12">
<h1 className="text-2xl font-bold tracking-tight">Editor</h1>
<p className="mt-1 text-sm text-muted-foreground">
Five ways to reach the model without leaving the sentence you were
writing — and the two surfaces that say where its answer came from.
</p>
</section>
<AiSlashMenu />
<AiInlineSuggestion />
<SelectionAiToolbar />
<AiActionMenu />
<AiInspectorPanel />
{/* Provenance and accountability, for the person about to publish. */}
<GroundingSplit />
<HumanOversightSplit />
</main>
)
}
components/ — each block page has its own copy button.app/ai-editor-page.tsx. The imports already point at @/components/…, so they resolve with no edits.The component, its props, the design tokens it expects and the command that installs it — as one prompt. Paste it into Claude, Cursor, v0 or ChatGPT and what they build around it will match the rest of the catalog instead of inventing its own system.
The screen an app opens on: shell, header, KPI row, chart and activity feed — numbers first, shape second, changes last.
An agent working inside a real app: a transcript down the middle — reasoning, answer, then the one card that can act — with what it may read and what it noticed demoted to a rail.
A kanban board inside the app shell, with tabs that admit the board is one view among several and a header whose primary action is creating work.
The canonical CRUD list — header, toolbar, sortable table and pagination framed as one panel instead of three stacked cards.
A wizard for the session someone finishes and a checklist for the one they do not, with the only genuinely blocking step in front of both.
Built around the state a search page is in most of the time — empty — so recent queries lead and the facets sit in a sidebar that survives growing to twelve.
One run opened up in the order you would debug it — intent, actions, the retries that are invisible in both, then cost.
The queue, the same queue sorted by how long things have waited, and the policy that decided both — with the policy last so it is not scrolled past forever.
The blank thread answered with real starter prompts, the model picker with the three numbers the choice turns on, and the four surfaces a session grows into — attachments, branches, canvas, threads.
The three states an agent screen usually omits: the wait before the first token, the trace after something broke with retries costed in the open, and the diff it wants a person to accept.
Why the answer said what it said: what is connected, how stale it is, what the query was allowed to see, what fitted in the window and what fell out — plus the failed-search state a debugging tool is most needed in.
A table after real people have used it — columns reordered from the keyboard, a selection bar that asks page-or-all, subtotals that survive collapsing, cells edited in place, and the empty state its own filters produced.
The saved view and the comparison window first, because every figure below depends on both — then six charts, none of which load a charting library.
Thresholds with how often each has actually fired, what that produced, where it went — and the digest settings that exist because of the answer.
The flow between an empty product and a useful one, in the order it goes wrong: sample data offered with equal weight, a file, a URL, the column mapping everything stalls on, and an upload that survives the connection dropping.
A profile form where the role options state what they actually change, the avatar crop works from the keyboard, and a navigation guard names the fields it is about to throw away.
The furniture that is on every screen and belongs to none — navbar, mobile drawer, command palette, shortcut sheet, header search, scope switcher, and the two shapes a detail view takes without losing the list.
Backwards from now and forwards from it on one screen — a day-grouped timeline, a month grid where a busy Tuesday never changes the row height, and the state both are in on day one.
What the product looks like while onboarding happens: a tour step anchored to the control it explains, the skeletons of the slowest load it will ever do, and the checklist that outlives both.