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".
npx skills add https://github.com/popmechanic/VibesOS --skill launch
> 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:
░▒▓█▓▒░ ░▒▓████████▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓███████▓▒░░▒▓██████▓▒ ░▒▓█▓▒░░▒▓█▓▒░
░▒▓█▓▒░ ░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░ ░▒▓█▓▒░░▒▓█▓▒░
░▒▓█▓▒░ ░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░ ░▒▓████████▓▒░
░▒▓█▓▒░ ░▒▓████████▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░ ░▒▓█▓▒░░▒▓█▓▒░
░▒▓█▓▒░ ░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░░▒▓█▓▒░▒▓█▓▒░ ░▒▓█▓▒░░▒▓█▓▒░
░▒▓████████▓▒░▒▓█▓▒░░▒▓█▓▒░░▒▓██████▓▒░░▒▓█▓▒░░▒▓█▓▒░░▒▓██████▓▒░▒▓█▓▒░░▒▓█▓▒░
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.
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.
Launch the editor server:
VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
bun "$VIBES_ROOT/scripts/server.ts" --mode=editor --prompt "USER_PROMPT_HERE"
If no prompt was given, omit --prompt:
VIBES_ROOT="${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 the user chose Editor above, STOP. Do not read or execute anything below this line.
The editor UI handles setup, generation, preview, and deployment.
Check for cached auth. Editor mode handles auth in the browser, so this only runs for Terminal:
VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
bun --input-type=module -e "
import { readCachedTokens, isTokenExpired } from '$VIBES_ROOT/scripts/lib/cli-auth.js';
const tokens = readCachedTokens();
if (tokens && !isTokenExpired(tokens.expiresAt)) {
console.log('AUTH_OK');
} else {
console.log('AUTH_NEEDED');
}
"
AUTH_OK → proceed silently (do not mention auth)AUTH_NEEDED → ask: "To deploy apps, you'll need a Vibes account. Sign in now? (A browser window will open for Pocket ID — takes about 10 seconds.)" VIBES_ROOT="${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.
Only one check before collecting input:
| # | Check | Command | If True |
|---|-------|---------|---------|
| 1 | app.jsx exists | test -f app.jsx | Ask [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.
Ask [App prompt]: "What do you want to build? Describe the app you have in mind."
Store as appPrompt.
Ask [App name]: "What's the app name? (used for subdomain + database)"
If "Derive from prompt": generate URL-safe slug (lowercase, hyphens, max 30 chars). Store as appName.
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?"
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).
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:
Always ask the user — do not assume a default.
Ask [Billing]: "What billing mode for your SaaS?" AND [Title]: "App display title?"
Ask [Tagline]: "Describe your app's tagline (short punchy phrase)"
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).
Theme switching is handled by the live preview wrapper, not inside the app. The builder generates a single-theme layout. Set themeCount = 1.
VIBES_ROOT="${CLAUDE_PLUGIN_ROOT:-$(dirname "$(dirname "${CLAUDE_SKILL_DIR}")")}"
TeamCreate("launch-{appName}", "Full SaaS pipeline for {appName}")${CLAUDE_SKILL_DIR}/prompts/builder.md{appPrompt}, {appName}, {pluginRoot}{aiInstructions}: if openRouterKey is set, add rule about useAI hook (see vibes SKILL.md "AI Features"). If null, leave empty.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.
Blocked by T1. Check TaskList until builder completes.
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.
Ask [Preview]: "Want to preview the app before deploying?"
If yes: run bun "$VIBES_ROOT/scripts/server.ts" and tell the user to open http://localhost:3333. They can chat to iterate on the design and switch themes. When satisfied, stop the server and continue to 2.3.
Step A — Assemble:
VIBES_ROOT="${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:
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.
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?"
If broken, ask what's wrong and troubleshoot.
Send shutdown_request to "builder" (if spawned). Wait for response. Clean up team.
## 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
| Failure | Recovery |
|---------|----------|
| Builder generates invalid JSX | Read app.jsx, fix TS syntax / wrong hooks, re-save |
| Assembly has placeholders | Re-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+ min | SendMessage status check. If no response, take over task |
| Builder hardcodes DB name | Edit app.jsx: replace with useTenant() pattern before assembly |
Assess Kubernetes workloads and cluster configuration for AKS Automatic compatibility. Identifies incompatibilities, generates fixes, and guides migration from AKS Standard to AKS Automatic. WHEN: migrate to AKS Automatic, check AKS Automatic readiness, validate manifests for Automatic, assess cluster for Automatic compatibility, fix deployment for Automatic compatibility, identify AKS Automatic migration blockers, is my cluster ready for AKS Automatic.
Discovers available Azure OpenAI model capacity across regions and projects. Analyzes quota limits, compares availability, and recommends optimal deployment locations based on capacity requirements. USE FOR: find capacity, check quota, where can I deploy, capacity discovery, best region for capacity, multi-project capacity search, quota analysis, model availability, region comparison, check TPM availability. DO NOT USE FOR: actual deployment (hand off to preset or customize after discovery), quota increase requests (direct user to Azure Portal), listing existing deployments.
Interactive guided deployment flow for Azure OpenAI models with full customization control. Step-by-step selection of model version, SKU (GlobalStandard/Standard/ProvisionedManaged), capacity, RAI policy (content filter), and advanced options (dynamic quota, priority processing, spillover). USE FOR: custom deployment, customize model deployment, choose version, select SKU, set capacity, configure content filter, RAI policy, deployment options, detailed deployment, advanced deployment, PTU deployment, provisioned throughput. DO NOT USE FOR: quick deployment to optimal region (use preset).
Unified Azure OpenAI model deployment skill with intelligent intent-based routing. Handles quick preset deployments, fully customized deployments (version/SKU/capacity/RAI policy), and capacity discovery across regions and projects. USE FOR: deploy model, deploy gpt, create deployment, model deployment, deploy openai model, set up model, provision model, find capacity, check model availability, where can I deploy, best region for model, capacity analysis. DO NOT USE FOR: listing existing deployments (use foundry_models_deployments_list MCP tool), deleting deployments, agent creation (use agent/create), project creation (use project/create).
Intelligently deploys Azure OpenAI models to optimal regions by analyzing capacity across all available regions. Automatically checks current region first and shows alternatives if needed. USE FOR: quick deployment, optimal region, best region, automatic region selection, fast setup, multi-region capacity check, high availability deployment, deploy to best location. DO NOT USE FOR: custom SKU selection (use customize), specific version selection (use customize), custom capacity configuration (use customize), PTU deployments (use customize).
This skill should be used when working with LaminDB, an open-source data framework for biology that makes data queryable, traceable, reproducible, and FAIR. Use when managing biological datasets (scRNA-seq, spatial, flow cytometry, etc.), tracking computational workflows, curating and validating data with biological ontologies, building data lakehouses, or ensuring data lineage and reproducibility in biological research. Covers data management, annotation, ontologies (genes, cell types, diseases, tissues), schema validation, integrations with workflow managers (Nextflow, Snakemake) and MLOps platforms (W&B, MLflow), and deployment strategies.
Latch platform for bioinformatics workflows. Build pipelines with Latch SDK, @workflow/@task decorators, deploy serverless workflows, LatchFile/LatchDir, Nextflow/Snakemake integration.
Run Python code in the cloud with serverless containers, GPUs, and autoscaling. Use when deploying ML models, running batch processing jobs, scheduling compute-intensive tasks, or serving APIs that require GPU acceleration or dynamic scaling.
Take popmechanic/launch from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.