Use cases

An LLM that ships.

Vibe Studio Apps turn any LLM into a knowledge-worker collaborator that can distribute its own deliverables — diagrams, graphs, documents, workbooks, mindmaps, runnable Python scripts — as self-contained URLs. No accounts. No platforms. No file attachments. The recipient clicks, edits, reshares. The artifact circulates outside any walled garden, on the LLM's terms.

How the loop closes

A chatbot or assistant that only returns text is a dead-end. Vibe Studio Apps give it a return value that moves: a URL that opens a fully-rendered, editable artifact in any browser. Each touch of the artifact — by the LLM, by the human, by the recipient — produces a new URL. The artifact propagates without ever hitting an authentication wall.

Open the flywheel as a flowchart → watch the protocol diagram itself, then edit the boxes if you disagree. Or open the same idea as a knowledge graph to see where the links land in real workflows.

Without Vibe Studio AppsWith Vibe Studio Apps
LLM returns a wall of markdown LLM returns a link to an editable, rendered artifact
Human screenshots the chat to share it Human pastes one URL into Slack, email, CRM
Recipient can't edit without re-prompting Recipient opens, edits, exports, reshares
Each artifact lives inside one chat thread Each artifact lives in a URL — portable everywhere
Account / login required to view No account. Public-link safe. Browser-only.
Vendor lock-in: data lives in the platform Data lives in the URL hash. The user owns it.
Multi-LLM workflows need a shared store Artifacts pass between agents as URLs — no store needed

Where this fits

Knowledge-worker chatbot
Daily assistant
"Draft the Q3 board memo." → doc link. "Map our deal pipeline." → workbook link. "Sketch the new auth flow." → diagram link. The chatbot stops being a text generator and starts being a colleague who hands you deliverables.
chat URL edit share
Open the example memo →
Slack / Teams bot replies
Inline collaboration
"@assistant, ERD for the new schema." The bot drops a flowchart-vibe link in the thread. Anyone opens it, edits, reshares the new link. The diagram lives in the conversation, not in a separate tool.
mention URL in thread team edits
Open the example sequence diagram →
Email-driven workflows
No attachments
LLM-drafted status updates, proposal docs, comparison tables — sent as links in the email body. Recipient opens in browser. No file size limits. No "blocked attachment" emails. No "I can't open this xlsx on mobile."
email link opens anywhere
Open the example workbook →
Intelligence office / research desk
Briefings without a portal
Competitive briefs (doc-vibe), competitor entity maps (kg-vibe), market sizing models (xls-vibe), product positioning grids (flowchart-vibe quadrant). The research desk publishes by sending links — no SharePoint, no Notion seat, no DocSend.
research 4 links distribution list
Open the example entity map →
Sales engineering & proposals
Bespoke without overhead
Each prospect gets a custom architecture diagram, a comparison sheet, a tailored memo — all generated by the LLM, all sent as one paragraph of links. The prospect can edit the proposal in their browser, send the new link to procurement. Closing loop without an "open the file" friction.
prospect tailored links they redline
Open the example 3D scene →
Workflow automation
Pipeline output
An n8n / Zapier / cron pipeline runs an LLM step, gets back a payload, encodes a #vibe= URL, sends it to Slack/email/Linear comment/Jira ticket. The artifact becomes the side-effect, not a database row.
trigger LLM URL downstream
Open the example node graph →
Office administration
Recurring artifacts
Weekly status memo (doc), monthly P&L draft (xls), quarterly org chart (flowchart), annual stakeholder map (kg). The LLM emits the same five links every Monday morning. Approvers click, edit if needed, forward as their own work.
Monday 5 links approved
Open the example status memo →
Education & training
Materials as links
Instructor's LLM generates lecture notes (doc), course map (mindmap), prereq graph (kg), worksheet (xls). Students click links — no LMS account. They submit by editing and sharing the new link back. Asynchronous, account-free, browser-native.
lesson 4 links student returns
Open the example mindmap →
Compliance & audit-friendly
No server, nothing to subpoena
Every artifact is in the URL hash — never sent to a server, never logged, never persisted by a third party. Revoking access = stop sending the URL. Audit trail = the URLs you have. For regulated industries: no DPA needed, no data residency question.
URL only no server audit-clean
Open the example knowledge graph →
Multi-agent orchestration
Artifacts as IPC
A research agent emits a kg-vibe link. A drafting agent reads it, generates a doc-vibe link. A review agent reads the doc and emits a flowchart-vibe link with the recommended next steps. The vibe URL is the message bus between agents — no broker, no shared state.
research draft review ship
Open the agent-loop diagram →
Embedded in other apps
CRM, PM, helpdesk, finance
Your CRM's "AI summary" button returns a doc link instead of a popover. Your project tool's "AI status" returns a flowchart link of the dependency graph. Your finance app's "AI explain" returns a workbook link with the calculation. postMessage and window.XLSStudio.load() let host apps embed each studio directly.
host app LLM embedded studio
Open the example HTML artifact →
Personal AI memory
Your stuff in your URLs
Your assistant remembers your projects via the links it's emitted. Save them in a notes app, an email folder, a bookmark list — they're complete. Click any one to land back where you were, edit, push forward. No subscription. No data lock-in.
bookmark resume extend
Open the example mindmap →
Runnable analysis, not screenshots
Computation as a link
When the user asks for an analysis, return the model, not the answer. A py-vibe link opens Python Studio with CPython 3.12 + numpy / pandas / matplotlib in their browser — script runs on open, plots render inline. They tweak the assumptions, re-run, reshare the new link. No "can you re-run that with 2026 numbers?" — they do it themselves.
analysis py-vibe link user tweaks reruns
Open the example analysis →

Media Studio — one example per kind

Every other studio renders structure (diagrams, graphs, sheets, code). Media Studio renders the raw bytes — images, audio, video, PDFs, fonts, 3D models, plain text, markdown — either inline (base64 in the URL hash, up to ~32 KB in Standard mode) or by pointer (any HTTPS src: URL — a gist raw, a GitHub raw, your own CDN). The same studio, two transports, eight kinds of asset.

Image · inline SVG
image/svg+xml · ~600 B
A small vector mark for the Vibe protocol, base64-encoded into the URL hash. Sanitised through DOMPurify before injection — safe to render an arbitrary src:-less SVG.
942 B hash inline compact mode
Open the SVG icon →
Audio · OPUS chime
audio/opus · 1.2 s · 3.7 KB
A 1.2-second welcome chime synthesised offline, encoded at 24 kbps OPUS. Inline in the URL with a live waveform display under the audio control. Press play.
6.6 KB hash 1.2 s standard mode
Open the chime →
Video · postage-stamp WebM
video/webm · 1 s · 2.8 KB
A 1-second, 160×120, 12-fps VP9 clip — autoplay, looping, muted. Demonstrates that "video in a URL" is real for short loops at thumbnail sizes. Larger media goes to src: mode.
5.2 KB hash 160×120 standard mode
Open the postage-stamp loop →
Font · via src pointer
font/woff2 · src: jsDelivr
Inter 700 Latin subset, fetched from a public jsDelivr CDN via a src: URL. The hash itself is only ~375 B — the font lives elsewhere. Renders as a full type specimen page (display sizes, pangrams, glyph grid).
375 B hash remote src pointer
Open the type specimen →
Markdown · protocol haiku
text/markdown · 540 B
A short markdown document — heading, blockquote haiku, body text — rendered with marked + DOMPurify. The whole document is the URL.
1.1 KB hash rendered compact mode
Open the haiku →
Plain text · ASCII art
text/plain · 480 B
A monospace banner. Reminds us that the simplest media is text — and even ASCII art deserves a themed paper viewer rather than a chat code-block.
1.0 KB hash monospace compact mode
Open the ASCII art →

How big can the URL be? Media Studio offers four size modes: compact (≤ 6 KB · survives email/QR), standard (≤ 32 KB · Slack/Discord-safe, the default), large (≤ 256 KB · Chrome/Firefox only — postage-stamp video and small fonts fit), and unlimited (Chrome address bar only). When inline data exceeds the chosen mode, the studio's Publish to Gist button uploads the asset to a public GitHub gist with the user's PAT (stored only in localStorage) and rewrites the asset to a src: pointer with a tiny hash.

Design principles

1

The URL is the artifact

Whole payload in the hash fragment. No server lookup. No login. The link is sufficient — copy it, send it, file it, fork it.

2

Browser-native, offline-friendly

Each studio is a single HTML file. After first load it works on a plane. IndexedDB holds your local library; the URL holds the share.

3

No vendor in the loop

Atomic 47 Labs hosts the studio HTML. We never see your payloads — they live in URLs that pass between you and your recipients.

4

One JSON schema per studio

Skills are tiny: ~150 lines and one encoder. An LLM that knows the schema can hand you a working link from any input.

5

Native formats out

Every studio exports to standard formats — .md, .svg, .png, .xlsx, .html, .json. The link is the entry point; the export is the exit.

6

Themable + brand-flexible

Five themes ship by default. The CSS variables are public — fork for your brand, host on your own subdomain, point your LLM at it.

Start with one studio. Add the rest as you go.

Each studio is independent. Pick the one your team needs first, install the skill, and see what a link-based deliverable feels like compared to a wall-of-text reply.