Extend this cloned Helios example app — add new controls, scenes, knobs, image flows, or features on top of `@reactor-models/helios` (pinned to ^0.9.0) without breaking the patterns the existing code already uses. Covers the SDK's connection / events / messages model, the phase-based UI architecture, the state snapshot pattern, the atomic `setConditioning({ prompt, image })` command for image-to-video flows, mid-stream prompt switching, the curated scene library, and prompt design rules for smooth continuous video generation.
Installation
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Extend this cloned Helios example app — add new controls, scenes, knobs, image flows, or features on top of `@reactor-models/helios` (pinned to ^0.9.0) without breaking the patterns the existing code already uses. Covers the SDK's connection / events / messages model, the phase-based UI architecture, the state snapshot pattern, the atomic `setConditioning({ prompt, image })` command for image-to-video flows, mid-stream prompt switching, the curated scene library, and prompt design rules for smooth continuous video generation.
Building on this Helios app
You've cloned this folder and now you want to extend it — a new control, a new scene, a new model knob, a different UX. This guide explains the patterns the existing code uses and the rules to follow so your additions feel native instead of bolted on.
All the code referenced below already exists in this folder. Read this guide alongside the source.
What Helios actually is, in three sentences
Helios is a continuous, prompt-driven video model. Once it starts generating, it produces an unending stream of video on a single track (main_video) — there is no "request, get clip, end". You steer the scene mid-stream by changing the prompt (which the model picks up on the next chunk), or by changing the reference image (which retints the conditioning without restarting).
The frontend's job is to (a) start the generation, (b) keep the user steering it, and (c) gracefully reflect the model's state.
The four concepts you'll touch
Concept
What it is
Hook / API
Connection
The lifecycle of the model session (disconnected → connecting → waiting → ready)
Things the model sends BACK to you — including the all-important state snapshot.
useHeliosState((m) => …), useHeliosCommandError, useHeliosImageAccepted, etc.
Tracks
The model's video output, rendered as a live MediaStreamTrack.
<HeliosMainVideoView />
You almost never have to drop below this surface. If you find yourself reaching for @reactor-team/js-sdk directly, stop and re-read the typed hooks list — there's likely a typed hook you're missing. The one documented exception is the recording surface (see Capturing clips below), which is a base-SDK feature that the typed packages deliberately do not re-export.
The UI phase model
A real-time video session is not one screen — it's a state machine. This app maps that state machine to two visible UI phases, and each component decides for itself which phase it lives in:
Components self-hide via early returns on the snapshot. No orchestration logic in the parent — adding a new component means dropping it into the sidebar and putting the right early-return at its top.
When you add a new control, decide its phase first
Before writing a new component, decide which phase it belongs to:
Knob that primes a session (e.g. seed picker, super-resolution mode) → Setup phase. Early-return when generating.
Knob that adjusts the live scene (e.g. image-strength slider, schedule-prompt picker) → Live phase. Early-return when not generating.
Always-on (e.g. a stats panel) → no early return; just gate interactivity on status === "ready".
The status === "ready" half of these checks matters — without it, your component will render stale data from a previous session after a disconnect/reconnect.
What's intentionally not exposed (and where to add it)
The model offers more knobs than this app surfaces. Each one is straightforward to add — drop a new component into the matching phase and call the relevant typed method. The model itself supports:
Knob
Hook
Lives in
Notes
set_seed
useHelios().setSeed({ seed })
Setup (read once at start)
-1 = random.
set_sr_scale
useHelios().setSrScale({ sr_scale })
Live OR Setup
"off" | "2x" | "4x". Takes effect on the next chunk.
set_image_strength
useHelios().setImageStrength({ image_strength })
Live
0..1. Ignored when no image is set.
schedule_prompt
useHelios().schedulePrompt({ prompt, chunk })
Live
Queues a prompt at a specific future chunk.
Mid-stream image swap
Re-run the upload→setImage flow from a live-phase control
Live
set_image is a conditioning tweak; the prompt stays untouched.
For each: add a small component, drop it into the sidebar at the right phase, and the rest of the architecture (status gating, snapshot lifecycle, error surfacing) Just Works.
Auth — getJwt resolver + cacheable GET route
Two pieces work together: a Next.js GET route that mints (and caches) a session-scoped JWT server-side, and a getJwt resolver prop on <HeliosProvider> that calls it on every Coordinator HTTP hop.
getJwt, not jwtToken
@reactor-team/js-sdk ≥ 2.10.1 accepts a resolver anywhere it used to take a static string:
The example passes getJwt={fetchToken} to <HeliosProvider>. The SDK then re-invokes that function on every Coordinator HTTP call — POST /sessions/:id/uploads, GET /clips, ICE refresh, SDP renegotiation — so a token aging out mid-session can't 401 those hops. The legacy jwtToken="..." string prop still works but caches one value at construction time and breaks the moment that value expires.
The provider auto-stabilizes the resolver via useRef + useMemo, so the inline arrow form is safe — a parent re-render does not tear the session down. Do not wrap it in useCallback.
Clip surfaces (<ClipPlayer>, <ClipDownloadButton>, useClipDownload) auto-inherit getJwt via React context, so you do not pass it through SnapClip anymore — see Capturing clips.
The route — app/api/reactor/token/route.ts
Already implemented. You usually don't need to touch it, but here's why it works the way it does so you don't accidentally break it:
GET, not POST. Browsers don't cache POST responses. The route handler still POSTs to the Reactor API internally; the public route exposes itself as GET so the browser's HTTP cache can transparently serve repeat calls.
Cache-Control: private. Never public — JWTs are per-user and must not be shared across users by any CDN or proxy.
max-age derived from the server's expires_at, not a hardcoded number. The Reactor /tokens endpoint accepts an expires_after body and returns { jwt, expires_at }. The route uses expires_at to set the cache window so it always tracks what the server actually granted.
authorization_details scopes the token. The mint pins the JWT to this app's model with a bounded session budget (max_sessions): the browser's token can only create sessions for that one model and act on the sessions it created — everything else on the account answers 403. Never hand a browser an unscoped token; that is the API key's full user-level access in cookie-jar form.
Because the route is GET + cacheable, the getJwt resolver is also dumb on the wire — every Coordinator hop calls fetch("/api/reactor/token"), which 99% of the time comes back from the browser's HTTP cache without ever touching your server.
Wiring an identity-provider JWT instead (Clerk, Auth0, …)
If your app uses Clerk session tokens or any other short-TTL identity JWT (Clerk's getToken({ template: "reactor" }) ships with a default ~60s TTL), getJwt is the hook for that:
Returning "" suppresses the Authorization header entirely (use this for local-dev / unauthenticated paths). getJwt wins over jwtToken when both are passed.
Configuring autoConnect
<HeliosProvider> is initialized withoutautoConnect. The user clicks "Connect" so they see the disconnected → connecting → waiting → ready transitions. If you're shipping a polished product where you'd rather the connection happen on page load:
Just make sure your status indicator still surfaces the intermediate states (connecting, waiting for GPU) — sessions don't reach ready instantly, and you don't want users staring at an unexplained loading state.
The state snapshot — your UI's single source of truth
Helios emits a state message after every command and every completed chunk. Subscribe via useHeliosState, hold it in useState, and read fields off it. Don't aggregate chunk_complete, generation_started, generation_paused and try to reconstruct state yourself — the snapshot already contains everything.
True once start() has succeeded. Stays true through pause. Reset to false by reset(). This is the phase switch.
running
True while the model is actively producing frames. Equal to started && !paused.
paused
True after pause(), false again after resume().
current_prompt
The prompt currently driving generation. null (typed as unknown) before start(). Match this against the scene library to drive evolution UIs.
current_chunk / current_frame
Progress counters since the last reset / connect.
image_set
True if a reference image has been provided this session.
Clear the snapshot on disconnect
The SDK does not emit a final state message when the session ends. Without an explicit reset, the last snapshot from the previous session lingers in your component's state — so after a reconnect, your UI shows stale "we're still generating!" data until the new session's first state arrives.
Every component that holds a snapshot does this:
useEffect(() => {
if (status !== "ready") setSnapshot(null);
}, [status]);
When you add a new component that subscribes to useHeliosState, include this. Three lines, no abstraction needed.
Sending events — the typed methods
Every event Helios accepts has a typed wrapper on useHelios(). Always await them; they return a Promise that can reject.
const { setPrompt, setImage, setConditioning, start, pause, resume, reset } =
useHelios();
// Text-to-videoawaitsetPrompt({ prompt: "A serene mountain lake at sunrise…" });
awaitstart();
// Image-to-video (curated scene with both pieces known up front)awaitsetConditioning({ prompt: "A serene mountain lake…", image: ref });
awaitstart();
// Transportawaitpause();
awaitresume();
awaitreset();
setConditioning (@reactor-models/helios@^0.9.0) bundles a prompt and an image into a single data-channel message. Prefer it over separate setImage + setPrompt calls whenever both pieces of conditioning are known at the same time — see the next section for why.
Never reach for sendCommand("set_prompt", ...) when a typed method exists. You lose autocomplete and the param-name typo check.
Status-gate every interactive control
Sending an event when status !== "ready" is a no-op with a console warning. Surface this as disabled on the button so the user sees what's clickable:
On disconnect, the gate trips and your new control greys out automatically — exactly the same visual state as a freshly loaded, never-connected page.
Chaining commands — prefer atomic commands over acknowledgment waits
Events are fire-and-forget over a data channel. Each event the SDK sends is a separate message, and commands that carry uploads (like setImage) take longer to resolve on the model than commands that don't (like start). Naive chaining races the model:
// ❌ flickers on the first frameawaitsetImage({ image: ref });
awaitsetPrompt({ prompt: "..." });
awaitstart();
setImage carries a FileRef that the runtime resolves into a real upload before the model dispatches it. While that resolution is in flight, start (which carries no upload) sails past on the same data channel and the model dispatches it first. The first chunk is then generated with no image conditioning, the image lands a tick later, and the user sees a visible scene-shift on the second chunk.
setConditioning is the right answer when you know both up front
For "kick off an image-to-video session from a known prompt + image" — which is the entire startFromExample path in app/components/ImageStarter.tsx — use the atomic command:
That's the whole flow. No useRef, no one-shot resolver, no useHeliosImageAccepted wiring. Why this works:
setConditioning is one message on the wire. A single message can't be split or reordered, so the runtime can't insert start between the prompt and the image.
The model handles it as a single transaction: validate → decode → VAE-encode → commit. The prompt schedule and image latents are only mutated together, and only after both pieces succeed. If anything fails (no prompt, no image, non-image MIME, undecodable bytes), the model emits command_error and mutates nothing — there's no partial state to recover from.
After it lands, the model emits prompt_accepted, image_accepted, conditions_ready(True, True), and a fresh state snapshot in one go. By the time the next message on the queue (start) is dispatched, conditioning is fully ready.
Use setConditioning whenever:
You're priming a fresh session with both pieces (curated scene launches, "load this preset" buttons).
The user has already typed a prompt and picked an image, and you want to commit them as one click.
When to fall back to setImage (image only) or setPrompt (prompt only)
The example only sends one piece at a time when there's a genuine human gap between them:
uploadCustomImage in ImageStarter.tsx calls setImage alone. The user uploads an image, then types a prompt in the textarea above, then clicks Start. By the time setPrompt + start fires from PromptComposer, the upload has long since been processed — no race.
EvolveScene.tsx calls setPrompt alone for mid-stream hot-swaps. The model is already running; there's no start to race against.
Rule of thumb: if you're about to call start() and you need conditioning, use setConditioning for both pieces or setPrompt for just a prompt. Never chain setImage + setPrompt + start in the same function — that's the race case.
The general "wait for an ack before start()" pattern
Helios 0.9.0 covers the prompt + image case atomically, so the example doesn't need an acknowledgment wait anywhere today. The pattern below is preserved as a reference for any future slow command without an atomic wrapper (a hypothetical setControlNet, setLora, setReferenceAudio, …). Subscribe to that command's *_accepted message and gate start() on it:
// Reference pattern. Not used by the current example —// `setConditioning` covers the only chained case we have.const readyRef = useRef<(() =>void) | null>(null);
useHeliosSomethingAccepted(() => {
if (readyRef.current) {
readyRef.current();
readyRef.current = null;
}
});
asyncfunctionstartWithSlowStep(params) {
// Park the resolver BEFORE sending the command — registering it// afterwards would race the model's response.const ready = newPromise<void>((resolve) => {
readyRef.current = resolve;
});
awaitsetSomethingSlow(params);
await ready; // ← gate start() on the ackawaitstart();
}
Prefer specific *_accepted messages over the broader conditions_ready — the latter fires once per conditioning command with partial flags, which makes predicate matching finicky.
The text-only path (setPrompt → start) never needs an ack wait — prompt updates are instant.
Receiving messages — the typed hooks
@reactor-models/helios ships one typed subscription hook per message:
Hook
Purpose
useHeliosState(handler)
The state snapshot. Use this; almost everything you need is here.
useHeliosCommandError(handler)
A command was rejected (bad preconditions, bad input). Render this somewhere visible.
useHeliosImageAccepted(handler)
An uploaded image was successfully processed. Not required for the curated image-to-video flow — setConditioning already commits both pieces atomically before start can be dispatched. Useful for "image uploaded ✓" toasts on setImage-only flows, or as the gate for a hypothetical future slow command that doesn't have an atomic wrapper.
useHeliosPromptAccepted(handler)
A prompt was queued. Useful for toast notifications.
useHeliosChunkComplete(handler)
One chunk finished generating. Useful for progress sounds, telemetry.
useHeliosGenerationStarted / Paused / Resumed
Lifecycle transitions. Useful for one-shot reactions (toasts, sounds), but don't aggregate these into your own state — read the snapshot instead.
useHeliosMessage(handler)
Catch-all over the typed discriminated union. Useful for devtools / logging.
useHeliosConditionsReady(handler)
Fires after each conditioning command with has_prompt / has_image flags. Lower-level than image_accepted; prefer the specific hooks.
Always surface command_error
app/components/CommandError.tsx already does this. The pattern:
"use client";
import { useState } from"react";
import { useHeliosCommandError, useHeliosState } from"@reactor-models/helios";
exportfunctionCommandError() {
const [err, setErr] = useState<{ command: string; reason: string } | null>(
null,
);
useHeliosCommandError((m) =>setErr({ command: m.command, reason: m.reason }),
);
useHeliosState(() =>setErr(null)); // any state update means the user moved onif (!err) returnnull;
return (
<divclassName="error">
{err.command} failed: {err.reason}
</div>
);
}
When you add a new event method, this will surface its failures automatically — no changes needed.
Image-to-video flow
Two patterns live in app/components/ImageStarter.tsx, one per launch path:
Curated scene — prompt + image known at the same time
Get bytes (fetch(url).then(r => r.blob()) for a curated image, or e.target.files[0] from <input type="file">).
await uploadFile(blob) → returns a FileRef.
await setConditioning({ prompt, image: ref }) — atomic. The SDK lifts the FileRef out of the params into an uploads envelope automatically; you treat image: ref as a regular field. Both pieces of conditioning are validated, decoded, VAE-encoded, and committed as one transaction. On failure the model emits command_error and leaves state untouched.
await start(). No acknowledgment wait needed — setConditioning lands before start reaches the model.
Custom upload — user picks an image, types a prompt later
const ref = awaituploadFile(file);
awaitsetImage({ image: ref });
// …user types in <PromptComposer>, which fires setPrompt + start
The user types in the textarea above and clicks Start. PromptComposer fires setPrompt + start. By that point the human delay has long covered the model's image processing, so there's no race — and no acknowledgment wait — in this path either.
Mid-stream image swap
Once started === true, calling setImage is a conditioning tweak, not a restart. Leave the prompt alone — overwriting the user's active prompt is destructive. (Mid-stream image swap UIs aren't shipped in this example; add one as a live-phase component if you want it.)
The scene library — one source, three surfaces
All suggested prompts live in app/lib/prompts.ts. Each scene is self-contained:
exportinterfacePrompt {
title: string; // short headline for the button labeltext: string; // full paragraph sent to the model
}
exportinterfaceScene {
id: string;
label: string;
initial: Prompt;
evolutions: ReadonlyArray<Prompt>;
imageUrl?: string; // present on image-backed scenes
}
exportconstSCENES: ReadonlyArray<Scene> = [
/* ... */
];
exportconstTEXT_SCENES = SCENES.filter((s) => !s.imageUrl);
exportconstIMAGE_SCENES = SCENES.filter(
(s): s is Scene & { imageUrl: string } => !!s.imageUrl,
);
exportfunctionfindSceneForPrompt(prompt: string | null | undefined,
): Scene | null {
if (!prompt) returnnull;
return (
SCENES.find(
(s) =>
s.initial.text === prompt ||
s.evolutions.some((e) => e.text === prompt),
) ?? null
);
}
The library feeds three surfaces:
PromptComposer reads TEXT_SCENES → renders initial of each as a clickable card.
ImageStarter reads IMAGE_SCENES → renders thumbnails (and ships the image bytes from /public/images/).
EvolveScene calls findSceneForPrompt(snapshot.current_prompt) → renders that scene's evolutions as one-click hot-swaps.
Adding a new scene = one entry in SCENES. No component changes. If the new scene has an imageUrl, drop the image bytes into public/images/ and reference it.
Prompts must be full paragraphs with explicit visual continuity
This is the most underrated part of building a real-time video frontend, and the #1 reason scenes look choppy when they should be smooth.
Each prompt is a paragraph, not a tagline. Describe the subject, the action, the environment, the lighting, AND the camera shot. Single-sentence prompts ("a cat chasing a butterfly") produce visually unstable output because the model has to invent everything else from scratch each chunk.
Each evolution must re-establish the subject before introducing the change. This is what lets the model hot-swap the prompt mid-stream without the scene visually resetting. Pattern:
Initial: "A majestic lion named Leo stands regally in the heart of a
dense jungle, embodying the essence of a king. Leo has a
golden mane that flows gracefully around his broad shoulders…
A medium close-up perspective emphasizing Leo's powerful
stance and the regal aura surrounding him."
Evolution: "Leo maintains his regal position on the rocky outcrop as
the humid jungle air settles around his broad shoulders. He
suddenly lowers his massive head to sniff a vibrant blue
butterfly that has fluttered near his nose…
A medium close-up perspective emphasizing Leo's powerful
stance and the regal aura surrounding him."
The camera shot, subject identity, and environment are restated verbatim. Only the action changes. That stability is what produces smooth on-screen continuity.
UI for long prompts
Render only the short title as the button label, with line-clamp-2 over the dim text underneath. The full text reaches the model on click; the truncation is purely visual:
This is Helios's signature capability. Once started === true, calling setPrompt({ prompt }) is a hot-swap on the next chunk — no restart, no start() again. The scene continues from exactly where it was, with the new prompt influencing subsequent frames.
app/components/EvolveScene.tsx is the canonical pattern: find the active scene by matching snapshot.current_prompt against the library, render its evolutions as click handlers that call setPrompt directly:
Notice: no start() in the click handler. The model is already generating; we're just swapping the prompt. The snapshot updates on the next chunk, the active evolution is highlighted/disabled, and the model picks up the new prompt smoothly.
If you build a new control that mutates the live scene (e.g. a slider for image strength), follow the same pattern — call the typed method, don't touch start().
Capturing clips
The Reactor base SDK exposes a recording surface that works for every model: ask for the last N seconds of the live stream, get back a Clip, and either preview it with <ClipPlayer> or download it with <ClipDownloadButton>. The model SDK does not own this — it lives on @reactor-team/js-sdk because it is the same call for Helios, Lingbot, and every future model with recording enabled.
The example ships a drop-in app/components/SnapClip.tsx panel that wires this together: a "Capture" button that calls requestClip(durationSeconds) off the store, opens a modal with the SDK's preview player, and offers an MP4 download. It is model-agnostic — the same file ships unchanged in every example.
When to reach for @reactor-team/js-sdk directly
The default rule still applies: do everything via @reactor-models/helios. But the typed package only re-exports model-specific surface (events, messages, the typed provider/hook). The recording surface is base-SDK only, so for that one feature you import directly:
When you scaffold a new component, ask: "Does this depend on Helios-specific events, messages, or commands?" If yes → typed package only. If no, and it would work the same on any model (recording, generic stats, generic connection state) → @reactor-team/js-sdk is fine.
The pattern
SnapClip is small enough to read in one go. The shape that matters:
Destructure the recording action off the store.useReactor((s) => s.requestClip) is the canonical accessor as of @reactor-team/js-sdk ≥ 2.11.1 — requestClip, requestRecording, and downloadClipAsFile are first-class actions alongside connect / disconnect / uploadFile. No s.internal.reactor indirection.
Gate on connection status.useReactor((s) => s.status) — return null when status is not "ready", so the panel disappears on disconnect just like every other live-only control.
Catch RecordingError. Recording can fail with typed reasons (DISCONNECTED, RECORDER_DISABLED, INVALID_DURATION, REQUEST_TIMEOUT). Surface them inline like CommandError does.
Compose <ClipPlayer> + <ClipDownloadButton> in a modal, route their errors through callbacks. Both accept onError (and <ClipDownloadButton> also accepts onSuccess(blob)); SnapClip threads them into the same inline error line that requestClip failures use. The SDK's components stay usable after disconnect, so the modal keeps working if the session ends mid-preview.
No getJwt plumbing. Both clip components auto-inherit the resolver from <HeliosProvider getJwt={…}> via React context (@reactor-team/js-sdk ≥ 2.10.1). That is the single source of truth for auth in this app — SnapClip doesn't need to know about the JWT route at all.
The portal gotcha (Sonner toasts, headless modals)
The context-inheritance only works for components rendered inside the provider subtree. SnapClip's modal is a normal child of the panel, so it inherits the resolver fine.
The trap is rendering clip UI through a React portal whose host lives outside<HeliosProvider> — most commonly a Sonner <Toaster /> mounted in app/layout.tsx as a sibling of {children}. The custom-toast tree has no ReactorContext in scope, the fallback returns undefined, and Coordinator answers the clip download with:
{"error":"Missing Authorization header"}
Fix: capture the resolver imperatively inside the provider subtree, then thread it down as an explicit prop. The resolver outlives disconnect() by design, so the toast keeps minting fresh tokens even after the session ends.
functionHandlerInsideProvider() {
// `requestClip` is a top-level store action; `internal.reactor`// stays in the escape-hatch slot for `getJwtResolver()` because// that one isn't lifted onto the store surface.const { requestClip, reactor } = useReactor((s) => ({
requestClip: s.requestClip,
reactor: s.internal.reactor,
}));
constonSnap = async () => {
const clip = awaitrequestClip(30);
// Captured here — works because we're inside the provider.// The closure carries it across Sonner's portal boundary.const getJwt = reactor.getJwtResolver();
toast.custom(() =><ClipReadyToastclip={clip}getJwt={getJwt} />);
};
}
hls.js is an optional peer
<ClipPlayer> plays HLS natively on Safari/iOS. On Chrome / Firefox / Edge it dynamically imports hls.js — which is why the example declares it as a direct dep (hls.js@^1.6.0). If hls.js isn't installed, the player surfaces an inline error and downloads still work; the dep keeps the preview path functional for the majority of users.
Extending
The component takes optional props for durationSeconds (default 10), filename, and label. Most extensions are one prop:
For multi-clip galleries, store an array of Clip instead of a single one, render a thumbnail per entry, and pass each to <ClipPlayer> / <ClipDownloadButton> on click. The headless useClipDownload hook is what to use if you want a custom progress UI instead of the default button.
Full-session recordings
requestRecording() (no args, also on the store) grabs everything from the start of recording up to now, instead of a trailing window. Same Clip shape, longer manifest, larger MP4. Swap the call inside SnapClip if you want a "Save the whole session" button instead.
Clips are short-lived
The URL on a Clip expires after a few minutes. Do not store Clip objects long-term, and do not hand clip.playlistUrl to your users for sharing. If you want a permanent link, download the MP4 (via <ClipDownloadButton> or downloadClipAsFile(clip, null) for a Blob) and host the result yourself.
Clip downloads are social-media-ready (@reactor-team/js-sdk ≥ 2.10.1)
Behind the existing <ClipDownloadButton> / reactor.downloadClipAsFile() API the SDK now remuxes the fragmented MP4 the runtime ships into a flat MP4 with start_time=0 and faststart layout, using mp4box as a bundled runtime dep. The transformation is ffmpeg -c copy style — no decode, no re-encode, the H.264 / AAC bitstream is bit-identical. The Blob you get back uploads cleanly to Twitter, Instagram, TikTok, YouTube; opens with start_time=0 in QuickLook; and on the rare parse failure falls back silently to the previous fragmented bytes (logged via console.warn), so the download never fails outright. No knob, no API change — you get the right behaviour by being on 2.10.1+.
Brand alignment — design tokens, not components
The app pulls Reactor's design tokens (fonts + brand colors) from @reactor-team/ui, but does not import its React components. Components ship interactive hooks under the hood — importing <Button> or <CodeSnippet> would force the consuming file into "use client" land. Design tokens have none of that baggage.
Use bg-brand, text-brand, font-mono etc. as plain Tailwind utilities — works in any component, server or client.
Reach for actual @reactor-team/ui components only when you need their behavior (e.g. a copy-on-click code block). Those usages are naturally Client Components anyway.
Common mistakes when extending
Reaching for @reactor-team/js-sdk directly. Everything Helios-specific is on @reactor-models/helios. If you find yourself reaching for useReactor((s) => s.internal.reactor) for a Helios event or message, re-read the typed hooks list above. The one allowed exception is the recording surface — see Capturing clips. And on 2.11.1+, even recording is a top-level store action (s.requestClip / s.requestRecording / s.downloadClipAsFile); reach for s.internal.reactor only for the few surfaces that aren't lifted yet (getJwtResolver(), raw runtimeMessage subscriptions).
Aggregating events to reconstruct state. Subscribe to useHeliosState and read fields off the snapshot. Stop folding chunk_complete + generation_started + generation_paused into your own boolean flags.
Chaining setImage + setPrompt + start instead of using setConditioning. Separate commands can be reordered on the wire — start slips past the still-resolving image upload and the first chunk renders with no image conditioning. When both pieces are known up front, use setConditioning({ prompt, image }) for a single atomic message. Only fall back to setImage alone when the prompt arrives later from a different user action (the custom-upload flow).
Overwriting the prompt on custom-image upload.set_image is a conditioning tweak. Leave the user's prompt alone — let them type one in the textarea if they need to.
Forgetting to clear the snapshot on disconnect. The next session's UI will show stale state. Three lines of useEffect in any component that holds a snapshot.
if (snapshot?.started) return null without a status check. After disconnect, snapshot.started may still be true (stale until the effect clears it). Always gate on status === "ready" too.
Connecting from a useEffect in your own component. The Provider owns connection lifecycle. Don't fight it; configure it via the connectOptions prop instead.
Importing @reactor-team/ui components into a Server Component. They use hooks internally. Either keep them in Client Components, or use the design tokens via CSS vars instead.
Single-line prompts and inconsistent evolutions. The model needs paragraph-length prompts with explicit visual continuity. Short prompts produce choppy output; mismatched continuity makes scenes visibly reset.
A new custom hook for one component. Inline the pattern first. Extract when you have three call sites of the same logic.
Breaking the file-per-phase rule. A single component that does setup AND live work is hard to read. Split it.
Checklist for new components
Before merging a new control or feature:
Decided which phase it lives in (Setup, Live, or always-on)
Early-return at the top matches that phase (status === "ready" && snapshot?.started to hide in live, etc.)
If it subscribes to useHeliosState, it clears on disconnect via useEffect
All interactive controls gate disabled on status === "ready"
All event method calls use the typed wrappers (setPrompt, not sendCommand("set_prompt", …)) and are awaited
If priming a session with prompt + image, uses setConditioning({ prompt, image }) (atomic). If chaining a different slow conditioning step before start() that doesn't have an atomic wrapper, awaits its specific *_accepted message via a one-shot ref resolver
Renders command_error somewhere visible (the existing CommandError component handles this automatically — don't suppress it)
New prompts added to app/lib/prompts.ts are full paragraphs with continuity between initial and evolutions
Brand colors via Tailwind utilities (bg-brand, text-brand), not hardcoded hex
No imports from @reactor-team/js-sdk or @reactor-team/ui React components unless absolutely required (recording surface is the documented exception — see Capturing clips)
