One component is useful. An assembly is powerful. These recipes show how to wire multiple manifests together to build complete features — ones your LLM can assemble from a single instruction.
A real-time voice assistant: captures the microphone, transcribes speech, generates a response with an LLM, and synthesizes it back to audio. Four manifests, zero glue code.
Full Stripe checkout with email confirmation: creates the payment session, listens to the webhook, and automatically sends a receipt.
Plug-and-play Retrieval-Augmented Generation: embed text chunks in-memory, search by semantic similarity, and generate a context-aware response with an LLM.
OAuth authentication with persistent sessions and abuse protection. The essential trio for any public-facing API.
Broadcast notifications across Slack, email, and mobile push from a single event bus. Add or remove channels without touching business logic.
Scrape a website, extract structured data with an LLM, and export the results as a clean CSV. A three-step data pipeline with no manual parsing.
An autonomous AI agent that can search the web, query a database, and send Slack messages — all through LLM function calling. Define tools once, let the model decide when to use them.
End-to-end subscription billing: manage recurring payments, generate PDF invoices, render branded email templates, and send them automatically. Event-driven — invoice is generated and emailed on every successful payment.
Production monitoring stack: health checks on dependencies, uptime monitoring on public endpoints, centralized error tracking — all funneled through a notification hub that fans out to Slack, email, and push.
Ingest RSS feeds, classify articles by topic with AI, render prompts from templates, generate summaries with an LLM, and validate outputs with guardrails. A full content processing chain.
Complete authentication pipeline: SAML SSO login, JWT token issuance, persistent sessions, and role-based access control. One flow from login to permission check.
Transcode uploaded videos to web-friendly formats, upload to S3, and notify the team via a hub. Event-driven — each step triggers the next automatically.
Track user events, segment users by behavior, run A/B tests with variant assignment, analyze conversion funnels, and export reports to CSV. A complete product analytics stack.
A conversational chatbot that follows scripted flows, escalates to an LLM when stuck, validates all AI responses with guardrails, and notifies support on Slack when it can't resolve the issue.
Share your component assembly on Discord and we'll add it to the cookbook.
Submit a recipe