Build, preview,
and deploy—one verified step at a time.
Practical paths for turning an idea, repository, template, or AI-generated codebase into software you can inspect and keep shipping.
- 11
- Practical guides
- 5
- Production stacks
- 01
- Release contract
- 01Sourcecontrolled
- 02Buildreproducible
- 03Previewinspectable
- 04Productionapproved
Choose your route
Start where your code is.
Both paths end with reviewable source and a production release. Pick the guide that matches what you have today.
Route 01
Start with an empty folder
Build and deploy a Next.js app
Create the app, add health and search metadata routes, define adios.yaml, verify production, and push the working source.
- Input
- Empty folder
- Outcome
- Live, verified application
Route 02
Start with a product brief
Build your first app with AI
Start from a working template, guide the built-in AI agent in plain language, inspect a preview, and publish your first live app.
- Input
- Plain-language idea
- Outcome
- Reviewed source + live URL
The Adios release loop
Learn the system, not a bag of tricks.
Source, runtime configuration, preview evidence, and production approval form one repeatable loop. Learn it once, then apply it to every stack.
- 01
Workspace
Keep the source in the loop
Use a source-backed workspace so every AI-generated change stays reviewable, testable, and Git-controlled.
/ AI coding/ Git/ Workspaces - 02
Contract
Describe how the app runs
Define builds, ports, health checks, secrets, managed resources, and routing beside the application source.
/ adios.yaml/ Manifest/ Runtime - 03
Release
Move AI-generated code to production
Build a repeatable path through preview, logs, secrets, Git, approval, and a production deployment.
/ Preview/ Logs/ Deployment - 04
Extend
Create and deploy an MCP server
Build a custom MCP server, develop it locally, store its secrets safely, and run it as an Adios application.
/ MCP/ Secrets/ Deployment
Build something real
Five stacks. One deployment contract.
Follow complete build references for real-time systems, AI search, APIs, documentation, and data-backed services.
Real-time analytics dashboard
Stream live visitor activity with a long-running Node.js runtime and managed data services.
$ socket.emit("visitor:update")AI semantic search engine
Serve Gemini embeddings and vector search without cold-start-heavy serverless glue.
$ POST /search → cosine_rank()Global Go webhook processor
Ingest webhooks with a compact Go API and persistent SQL connection pools.
$ go run ./cmd/workerModern PHP docs site
Combine PHP, MySQL, Markdown rendering, and cache headers for fast dynamic documentation.
$ php -S 0.0.0.0:8080.NET crypto price ticker
Buffer price updates with Redis and deploy a .NET Minimal API without IIS or Azure setup.
$ dotnet run --urls :8080Quick references
Ready when the source is