hz README

The project's README from github.com/danielvaughn/hz. As one markdown file: README.md.

Huzzah

An experimental interface for building software by editing persistent pseudocode and synchronizing it with an AI-generated implementation.

setup

You need Node.js 22.19 or newer and access to a model provider supported by Pi.

  1. git clone https://github.com/danielvaughn/hz.git
  2. cd hz
  3. npm install
  4. Configure Pi using one of the options below.
  5. npm run dev

Open the url (should run on http://localhost:5173)

configure Pi

Huzzah uses Pi's provider, credential, and model configuration. It does not require a particular model vendor.

authenticate with an API key

Set an environment variable supported by your provider before starting Huzzah. For example:

export ANTHROPIC_API_KEY=sk-ant-...

Pi also recognizes credentials for OpenAI, Google, Azure OpenAI, Amazon Bedrock, and other supported providers.

authenticate with a subscription

Run Pi through the web app's local installation:

npx --workspace=poc pi

Enter /login, choose a provider, and complete its authentication flow. Exit Pi when authentication is complete, then start the web app with npm run dev.

choose a model

Set Pi's default provider and model in ~/.pi/agent/settings.json:

{
  "defaultProvider": "anthropic",
  "defaultModel": "claude-sonnet-4-5"
}

Use model IDs available to your configured provider. If no default is set, Pi falls back to the first authenticated model it finds. Huzzah currently uses Pi's low thinking level for reconciliation and semantic highlighting regardless of defaultThinkingLevel.

Pi also supports custom providers and local models such as Ollama, LM Studio, and vLLM through ~/.pi/agent/models.json.

Do not paste secrets or private source into Huzzah. Specs and current generated source are sent to your selected model provider, and accepted AI-generated JavaScript runs locally in a Web Worker. This is experimental containment, not a hostile-code sandbox.

ideas to get started

hz (danielvaughn/hz)

Searching Database…

201 ·GitHub
What is this?Huzzah ('hz') is an experimental editor where you write a persistent pseudocode spec and an AI keeps a JavaScript implementation continuously synchronized with it, so you build software by editing the spec instead of the code.
Is this a code editor like VS Code, or something different?It is a different paradigm: you edit a pseudocode spec and the AI regenerates and reconciles the matching JavaScript implementation, which runs locally in a Web Worker rather than being hand-written.
How do I get it running?Clone the repo, run npm install, configure Pi's provider/credential settings, then run npm run dev to launch the web app in your browser.
What do I need before trying it?You need Node.js 22.19 or newer and access to any model provider supported by Pi — authenticate via an API key (OpenAI, Google, Azure OpenAI, Amazon Bedrock, etc.), a subscription login, or a local model like Ollama, LM Studio, or vLLM.
Is the AI-generated code safely sandboxed?Not fully — running generated code in a Web Worker is described as experimental containment, not real isolation from hostile code, so you shouldn't rely on it as a security boundary.
Who is this meant for?It's aimed at developers curious about an AI-mediated, spec-first way of writing software rather than traditional hand-coding.
Does it run in the browser, or is there a desktop app too?It runs primarily as a web app opened via npm run dev, and the repo also includes an Electron build (dev:electron, build:electron) for a desktop version.
Does my spec or code get sent anywhere?Yes — your specs and the currently generated source are sent to whichever model provider you've configured, so the README warns against pasting secrets or private source into Huzzah.
What is Pi?Pi is the provider, credential, and model configuration layer Huzzah relies on — it abstracts away the specific AI vendor, letting Huzzah work with many different providers and local models.
What kind of projects does the README suggest trying?Small, well-scoped logic apps like a shopping cart with discount rules, a turn-based combat system, a budget tracker, a rate limiter, and a word game — good test cases for the spec-to-code workflow.
Do I need a specific AI model or vendor?No — Huzzah doesn't require a particular model vendor; it just needs one supported by Pi, and falls back to the first authenticated model if you haven't set a default.
Can I use a local model instead of a cloud API?Yes, Pi supports custom providers and local models such as Ollama, LM Studio, and vLLM configured through ~/.pi/agent/models.json.
Size
0.3 MB
Stars
201
Updated
CommitAug 23
Age
2mo
Commits · 90 Days