Edit Bay Stream
Lovable AI categorizes your stream-of-consciousness notes so editors get organized post-production tickets; ElevenLabs captures it.
Realtime Speech-to-Text· live transcription
Section · Voice
full primer →The primitive.
Filmmakers speak; the ElevenLabs scribe model streams live captions back as they go, finalizing each line during the natural pauses in post-production review.
Why this primitiveStream-of-consciousness editing notes require continuous capture to preserve the editor's rapid-fire critical flow.
Kernel
the @elevenlabs/react useScribe hook on scribe_v2_realtime — server mints a single-use scribe token, client streams microphone audio and receives partial + committed transcripts with word timestamps
Drives the UI as
a live caption strip that types itself as the user speaks, with finalized lines saved as the user pauses
Required key.
ELEVENLABS_API_KEY
Single key for TTS, voice agents, scribe, music and SFX. Free tier covers a hackathon weekend.
open ↗Add this in your Lovable project under Settings → Secrets before pasting the prompt below.
Appendix · Mega-prompt
The build prompt.
budget · 1 message
Paste into a fresh Lovable project. Make sure the key above is set first. read the build strategy →
Build "Edit Bay Stream" as a ONE-SHOT Lovable build. The participant has only
5 credits — this single message must produce a working demo with no follow-ups.
Single-page TanStack Start app. Cut scope ruthlessly.
CONCEPT
Lovable AI categorizes your stream-of-consciousness notes so editors get organized post-production tickets; ElevenLabs captures it.
Discipline: Filmmaking & Animation (post-production review).
Recipe: Lovable AI brain + Realtime Speech-to-Text (live transcription) as the voice surface.
Why this voice surface: Stream-of-consciousness editing notes require continuous capture to preserve the editor's rapid-fire critical flow.
LOVABLE BUDGET (HARD CAP: ONE-SHOT, ~5 CREDITS TOTAL):
The participant has FIVE Lovable credits for the whole build. This prompt MUST
ship a working demo on the FIRST message with zero follow-ups. Engineer for that.
- ONE TanStack Start app, ONE route (`src/routes/index.tsx`). No extra pages, no auth, no nav.
- TWO TanStack server functions max: one for the Lovable AI call (the brain),
one for the ElevenLabs call (the voice). Fold them into one if the voice
primitive does not need server-side text generation.
- ONE client surface (a button, a mic, or a prompt box) wired to those server fns.
- NO database, NO Lovable Cloud, NO auth, NO file uploads, NO extra integrations.
- NO tests, NO docs pages, NO settings screens, NO theming toggles.
- Libraries: template defaults + `ai` + `@ai-sdk/openai-compatible` + `zod`,
plus (only if the voice kernel needs it) `@elevenlabs/react`. Nothing else.
- Keep the diff small enough to land in one build pass. If a feature is not on
screen in the user flow below, do not build it. Cut scope before adding scope.
STACK
- TanStack Start app, the index route only.
- Lovable AI Gateway (the brain) + ElevenLabs (the voice). All calls live
inside `createServerFn` handlers so both keys stay on the server.
- Client surface fits the kernel: a prompt box for TTS / music / SFX, a mic
for agents, a live caption strip for scribe. Render markdown from the brain
with `react-markdown` if you show the text on screen.
- Tailwind + shadcn. Editorial look: gold accent on a dark or warm-cream
background, generous type, one strong headline, one primary action.
- Footer renders: "Built during the Creative AI & Quantum Hackathon organised by StreetKode Fam during Indian Krump Festival 14".
BRAIN — Lovable AI Gateway (free for participants, no key prompt needed):
```ts
// src/lib/ai-gateway.server.ts
import { createOpenAICompatible } from "@ai-sdk/openai-compatible";
export function gateway() {
return createOpenAICompatible({
name: "lovable",
baseURL: "https://ai.gateway.lovable.dev/v1",
headers: {
"Lovable-API-Key": process.env.LOVABLE_API_KEY!,
"X-Lovable-AIG-SDK": "vercel-ai-sdk",
},
});
}
```
Default model: `google/gemini-3-flash-preview`. Use `generateText` (or
`streamText` for long output) from `ai`. Keep prompts and model calls inside
the server function — never call the gateway from client code.
SERVER FUNCTION (src/lib/scribe.functions.ts) — mint a scribe token:
```ts
import { createServerFn } from "@tanstack/react-start";
/** Built during the Creative AI & Quantum Hackathon organised by StreetKode Fam during Indian Krump Festival 14 */
export const mintScribeToken = createServerFn({ method: "POST" }).handler(async () => {
const r = await fetch(
"https://api.elevenlabs.io/v1/single-use-token/realtime_scribe",
{ method: "POST", headers: { "xi-api-key": process.env.ELEVENLABS_API_KEY! } },
);
const { token } = await r.json();
return { token };
});
```
OPTIONAL BRAIN PASS (src/lib/refine.functions.ts) — Lovable AI shapes the
raw transcript into something useful for post-production review:
```ts
import { createServerFn } from "@tanstack/react-start";
import { generateText } from "ai";
import { z } from "zod";
import { gateway } from "./ai-gateway.server";
/** Built during the Creative AI & Quantum Hackathon organised by StreetKode Fam during Indian Krump Festival 14 */
export const refine = createServerFn({ method: "POST" })
.inputValidator((d) => z.object({ transcript: z.string().min(1) }).parse(d))
.handler(async ({ data }) => {
const { text } = await generateText({
model: gateway()("google/gemini-3-flash-preview"),
system: `Turn the user's spoken post-production review notes into a clean, actionable result.`,
prompt: data.transcript,
});
return { text };
});
```
CLIENT (`bun add @elevenlabs/react` first):
```tsx
import { useScribe } from "@elevenlabs/react";
const scribe = useScribe({ modelId: "scribe_v2_realtime", commitStrategy: "vad" });
// connect with the minted token, render scribe.partialTranscript live, then
// call refine() with the joined committed transcripts when the user stops.
```
TRANSLATION — skip:
This kernel is non-linguistic (audio in or audio out), so do not add a
language selector or a translation pass. Keep the brain-to-voice path direct.
USER FLOW (the entire app — nothing else exists)
1. Land on the page; the headline previews what the demo does for post-production review.
2. The primary action (a live caption strip that types itself as the user speaks, with finalized lines saved as the user pauses) is one tap away; the rest of the layout supports it.
3. Lovable AI does the thinking, ElevenLabs handles the voice surface, and the
result (audio + any text) stays on screen so the user can retry or share.
KEYS — both already provided to participants for free:
1. `LOVABLE_API_KEY` (the AI brain). Auto-injected in every Lovable project.
Read it only on the server via `process.env.LOVABLE_API_KEY`. Never prefix
with `VITE_` and never expose to the client.
2. `ELEVENLABS_API_KEY` (the voice). Best path: ask Lovable to connect the
ElevenLabs standard connector — it syncs the key automatically and rotates
from the dashboard. Fallback: paste a key from
https://elevenlabs.io/app/settings/api-keys into Project Settings -> Secrets.
Read it only on the server, header `xi-api-key`, never prefix with `VITE_`.
CREDIT (must appear in UI footer AND as JSDoc on the server function):
Built during the Creative AI & Quantum Hackathon organised by StreetKode Fam during Indian Krump Festival 14
Market sizing.
TAM
$400B
global film and animation production
SAM
$18B
video editing and VFX software
SOM
$90M
freelance editor productivity tools
Indicative figures for hackathon pitches — refine with your own research before raising.
Adjacent entries.
narrative pitching
Storyboard Pitch Whisper
Lovable AI structures your live pitch so you finalize treatments faster; ElevenLabs captures it.
sound design spottingFoley Beatbox Catcher
Lovable AI translates your beatboxed sound effects so foley artists get precise cue sheets; ElevenLabs captures it.
voice acting directionDirector Emotion Logger
Lovable AI logs your emotional adjustments so voice actors get timestamped script notes; ElevenLabs captures it.
animation timingExposure Sheet Vocalizer
Lovable AI converts your vocalized frame counts so you build accurate exposure sheets; ElevenLabs captures it.