Launch logo

Launch

Community
popmechanic
launch

Self-contained SaaS pipeline — invoke directly, do not decompose. Generates a Vibes app, adds auth + billing, and deploys live. Uses Agent Teams to parallelize for maximum speed. Use when the user wants to build and ship a complete SaaS product in one step, or says "launch", "ship it", "build and deploy".

Overview

Publisherpopmechanic
RepositoryVibesOS
Skill namelaunch
Stars
134
Forks
6
Bundled files
2
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 2 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by popmechanic on GitHub. Read the source before you install it.

Installation

Install the Launch AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/popmechanic/VibesOS.git /tmp/VibesOS
mkdir -p .claude/skills
cp -r /tmp/VibesOS/skills/launch .claude/skills/launch
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Launch in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Launch on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Launch is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Plan mode: If you are planning work, this entire skill is ONE plan step: "Invoke /vibes:launch". Do not decompose the steps below into separate plan tasks.

Display this ASCII art immediately when starting:

░▒▓█▓▒░      ░▒▓████████▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓███████▓▒░░▒▓██████▓▒ ░▒▓█▓▒░░▒▓█▓▒░
░▒▓█▓▒░      ░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░     ░▒▓█▓▒░░▒▓█▓▒░
░▒▓█▓▒░      ░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░     ░▒▓████████▓▒░
░▒▓█▓▒░      ░▒▓████████▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░     ░▒▓█▓▒░░▒▓█▓▒░
░▒▓█▓▒░      ░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░     ░▒▓█▓▒░░▒▓█▓▒░
░▒▓████████▓▒░▒▓█▓▒░░▒▓█▓▒░░▒▓██████▓▒░░▒▓█▓▒░░▒▓█▓▒░░▒▓██████▓▒░▒▓█▓▒░░▒▓█▓▒░

Notation

Ask [Header]: "question" means call AskUserQuestion with that header and question. Options listed as bullets. User can always type custom via "Other".

For architecture context, see LAUNCH-REFERENCE.md in this directory.


FIRST: Terminal or Editor UI?

This is the very first question — ask before anything else. DO NOT check .env, credentials, or project state before asking this question. DO NOT invoke any other skill before asking this question. If Editor is chosen, skip ALL remaining steps — the editor handles everything.

Ask the user:

"How do you want to build? Editor (opens a browser UI with live preview, chat, and deploy button) or Terminal (I'll generate and deploy from here)?"

Present Editor as the first/recommended option.

  • If Editor: Start the editor server and END YOUR TURN. Do not ask any more questions. Do not continue to Phase 0 or any phase below. The editor UI handles the entire workflow — setup, generation, preview, deploy.

    Launch the editor server:

    bash
    VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
    bun "$VIBES_ROOT/scripts/server.ts" --mode=editor --prompt "USER_PROMPT_HERE"

    If no prompt was given, omit --prompt:

    bash
    VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
    bun "$VIBES_ROOT/scripts/server.ts" --mode=editor

    Tell the user: "Open http://localhost:3333 — the editor handles everything from here." Your job is done. Stop. Do not read further. Do not proceed to any phase below.

  • If Terminal: Continue with the workflow below.


⛔ EVERYTHING BELOW IS TERMINAL MODE ONLY

If the user chose Editor above, STOP. Do not read or execute anything below this line. The editor UI handles setup, generation, preview, and deployment.


Auth Check (Terminal mode only — silent unless needed)

Check for cached auth. Editor mode handles auth in the browser, so this only runs for Terminal:

bash
VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
bun --input-type=module -e "
import { readCachedTokens, isTokenExpired } from '$VIBES_ROOT/scripts/lib/cli-auth.js';
const tokens = readCachedTokens();
if (tokens && !isTokenExpired(tokens.expiresAt)) {
  console.log('AUTH_OK');
} else {
  console.log('AUTH_NEEDED');
}
"
  • If AUTH_OK → proceed silently (do not mention auth)
  • If AUTH_NEEDED → ask: "To deploy apps, you'll need a Vibes account. Sign in now? (A browser window will open for Pocket ID — takes about 10 seconds.)"
    • If yes:
      bash
      VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
      bun --input-type=module -e "
      import { getAccessToken } from '$VIBES_ROOT/scripts/lib/cli-auth.js';
      import { OIDC_AUTHORITY, OIDC_CLIENT_ID } from '$VIBES_ROOT/scripts/lib/auth-constants.js';
      const tokens = await getAccessToken({ authority: OIDC_AUTHORITY, clientId: OIDC_CLIENT_ID });
      if (tokens) console.log('Signed in successfully!');
      "
      Confirm success, then continue.
    • If no → proceed anyway (auth will be needed at deploy time)

Pre-Flight Check

Only one check before collecting input:

#CheckCommandIf True
1app.jsx existstest -f app.jsxAsk [Reuse]: "app.jsx exists. Reuse it or regenerate?" If reuse: skip T1.

Auth and deploy credentials are handled automatically — no user setup needed. On first deploy, a browser window opens for Pocket ID login. Tokens are cached at ~/.vibes/auth.json.

Phase 0: Collect Inputs

0.1 App Prompt

Ask [App prompt]: "What do you want to build? Describe the app you have in mind."

  • "Todo list" — A simple task manager with categories and due dates
  • "Photo gallery" — A shareable photo gallery with albums and captions
  • "Team dashboard" — A metrics and status dashboard for small teams

Store as appPrompt.

0.2 App Name

Ask [App name]: "What's the app name? (used for subdomain + database)"

  • "Derive from prompt" — Generate a URL-safe slug automatically
  • "Let me specify" — I'll type my own name

If "Derive from prompt": generate URL-safe slug (lowercase, hyphens, max 30 chars). Store as appName.

0.3 AI Features (conditional)

Scan appPrompt for AI keywords: "chatbot", "chat with AI", "summarize", "generate", "analyze", "AI-powered", "intelligent".

If detected: Ask [AI features]: "Does this app need AI features?"

  • "Yes — I have an OpenRouter key" — I'll paste my API key
  • "Yes — I need to get one" — I'll sign up at openrouter.ai
  • "No AI needed" — Skip AI capabilities

If yes: check grep OPENROUTER_API_KEY ~/.vibes/.env. If found, offer reuse (mask key). Otherwise collect via Ask and offer to cache to ~/.vibes/.env. Store as openRouterKey (or null if no AI).

0.4 SaaS Config

Factory config is collected here but applied later by the assembly script. Do NOT hand-implement SaaS logic — the factory assembly handles tenant routing, auth gating, and billing.

Choose billing mode based on monetization intent:

  • "off" (free) — all authenticated users get full access. Good for MVPs and internal tools.
  • "required" (subscription) — users must subscribe. Stripe billing integration is phase 2.

Always ask the user — do not assume a default.

Ask [Billing]: "What billing mode for your SaaS?" AND [Title]: "App display title?"

  • Billing: "Free (no billing)" or "Subscription required"
  • Title: "Derive from app name" or "Let me specify"

Ask [Tagline]: "Describe your app's tagline (short punchy phrase)"

  • "Generate one" — Create from app description
  • "Let me write it" — I'll provide it

When billing is "required": These fields appear on a pricing section visible to potential customers before signup. Optimize for marketing copy quality — benefit-driven language, not technical descriptions. Tagline = sales headline. Subtitle = value proposition ("why should I pay?"). Features = compelling benefit statements (3-5 items).

Repeat pattern for subtitle and features list (3-5 bullet points).

Store: billingMode ("off"/"required"), appTitle, tagline, subtitle, features (JSON array).

0.5 Theme Selection

Theme switching is handled by the live preview wrapper, not inside the app. The builder generates a single-theme layout. Set themeCount = 1.


Phase 1: Build

1.1 Setup

  1. Resolve plugin root — use this in all bash blocks:
    bash
    VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
  2. Create team: TeamCreate("launch-{appName}", "Full SaaS pipeline for {appName}")
  3. Create tasks: T1 (generate app.jsx), T3 (assembly), T4 (deploy), T5 (verify).

1.2 Spawn Builder (T1)

  1. Read ${CLAUDE_SKILL_DIR}/prompts/builder.md
  2. Substitute: {appPrompt}, {appName}, {pluginRoot}
  3. Set {aiInstructions}: if openRouterKey is set, add rule about useAI hook (see vibes SKILL.md "AI Features"). If null, leave empty.
  4. Spawn: Task tool, team_name="launch-{appName}", name="builder", subagent_type="general-purpose"

While the builder generates, SaaS config was already collected in Phase 0. Nothing else needs to happen in parallel — no credential collection, no manual setup.


Phase 2: Assembly & Deploy

Blocked by T1. Check TaskList until builder completes.

2.1 Verify Inputs

Confirm: app.jsx exists with valid JSX. All factory config values collected.

Scan app.jsx for builder mistakes (see LAUNCH-REFERENCE.md "Common Builder Mistakes"). Fix any found before proceeding.

2.2 Preview Before Deploy

Ask [Preview]: "Want to preview the app before deploying?"

  • "Yes — open live preview" — Start the preview server and iterate on the design
  • "No — deploy now" — Skip preview, go straight to deploy

If yes: run bun "$VIBES_ROOT/scripts/server.ts" and tell the user to open http://localhost:3333. They can chat to iterate on the design and switch themes. When satisfied, stop the server and continue to 2.3.

2.3 Assemble & Deploy

Step A — Assemble:

bash
VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
bun "$VIBES_ROOT/scripts/assemble-factory.js" app.jsx index.html \
  --app-name "{appName}" \
  --app-title "{appTitle}" \
  --domain "{appName}.vibes.diy" \
  --billing-mode "{billingMode}" \
  --tagline "{tagline}" \
  --subtitle "{subtitle}" \
  --features '{featuresJSON}' \
  --admin-ids '[]'

Step B — Validate: grep -c '__VITE_\|__OIDC_\|__APP_' index.html — must be 0.

Step C — Deploy:

bash
VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
bun "$VIBES_ROOT/scripts/deploy-cloudflare.js" \
  --name "{appName}" \
  --file index.html \
  {aiKeyFlag}

Where {aiKeyFlag} = --ai-key "{openRouterKey}" if set, omitted if null.

On first deploy, the CLI opens a browser for Pocket ID login. The user signs in once and tokens are cached for future deploys.


Phase 3: Verify & Cleanup

3.1 Verify

Ask [Verify]: "Your app is live! Open each URL and verify:\n\n- Landing: https://{appName}.vibes.diy\n- Tenant: https://{appName}.vibes.diy?subdomain=test\n\nDoes everything look right?"

  • "All working" — Everything loads correctly
  • "Something's broken" — Need to troubleshoot

If broken, ask what's wrong and troubleshoot.

3.2 Shutdown

Send shutdown_request to "builder" (if spawned). Wait for response. Clean up team.

3.3 Summary

## Launch Complete

**App**: {appTitle}
**URL**: https://{appName}.vibes.diy
**Billing**: {billingMode}

### What's deployed:
- Cloudflare Worker via Deploy API
- TinyBase sync via WebSocket + Durable Object (auto-provisioned)
- Pocket ID authentication with passkeys
- Subdomain-based multi-tenancy

### Next steps:
- Sign in at your app URL to become the first user
- Configure admin access via the app settings
- Set up Stripe billing if using subscription mode

Error Handling

FailureRecovery
Builder generates invalid JSXRead app.jsx, fix TS syntax / wrong hooks, re-save
Assembly has placeholdersRe-run assembly — auth constants are hardcoded, no .env needed
Deploy fails (auth)Delete ~/.vibes/auth.json and retry — browser login will re-open
Deploy fails (other)Check error output from deploy-cloudflare.js
Teammate silent 3+ minSendMessage status check. If no response, take over task
Builder hardcodes DB nameEdit app.jsx: replace with useTenant() pattern before assembly

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Launch AI skill do?

Self-contained SaaS pipeline — invoke directly, do not decompose. Generates a Vibes app, adds auth + billing, and deploys live. Uses Agent Teams to parallelize for maximum speed. Use when the user wants to build and ship a complete SaaS product in one step, or says "launch", "ship it", "build and deploy".

Why use Launch on TypingMind?

Because you install it once and use it with any model. Launch is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Launch in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/popmechanic/VibesOS/tree/main/skills/launch. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Launch?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Launch?

As many as you like. As long as a model supports skills, you can use Launch with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Launch AI skill free?

Yes. It is published on GitHub by popmechanic under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇