Zelyra 1.2.40 · in active development

Business apps
in one clear language.

Describe data, rules and web pages in one readable source. Zelyra connects the language, compiler, database tools, runtime and Studio so people and AI can build on the same project.

Quick Start → Explore Zelyra ↓ GitHub ↗

Source v1.2.40 GitHub ★ Star

🦀 Rust 2024

🐬 MariaDB Native

⚖️ Apache-2.0

customer.zyl example source

01table customers {
02id: Id primary auto
03name: String(100) required
04email: Email?
05}

model → rules → forms → Studio

◇

People + AI One readable source to work from

✓

Checks during development Detected issues in supported code, before deployment

01 / the idea

One language.
One connected workflow.

Zelyra brings the parts of application development into one project: describe a data model and business rules, connect supported database and web workflows, then inspect and edit the source in Studio. Clear syntax helps people understand the code; consistent structure and machine-readable compiler output help AI tools read and write it.

Understand the principle →

01 Business data

Data Model

Describe business data such as customers, articles and invoices in a visible model. Studio can inspect supported schemas without changing the database.

table customers { ... }

02 Logic

Types & Rules

Readable rules and explicit types give people and AI a shared source to work from. The compiler checks supported syntax and types and reports stable diagnostic codes.

type UserId = Id

03 Web

Web & Forms

Build supported customer and article forms, routes and CRUD from the same project. Studio keeps generated source reviewable and editable.

page "/customers" { ... }

04 SQL · charts planned

From data to insight

Zelyra supports documented SQL and database workflows. The Studio roadmap plans guided metrics such as new customers per month and invoice revenue over time.

New customers · by month

02 / why zelyra

⌘

Readable by people.
Ready for AI.

Consistent syntax, explicit types and structured project context help AI tools understand and write Zelyra. The compiler checks supported rules; human review remains part of the workflow.

name: String?
match name {

◇

One model.
Connected steps.

A project model can feed supported database, form, route and CRUD workflows. Studio helps inspect the relationships and review generated source.

table→form→API

↗

Less boilerplate.
More business logic.

Zelyra aims to replace repetitive glue code with model-driven workflows. Smaller changes can mean fewer files and less AI context; whether that reduces tokens and cost depends on the task and must be measured.

Open learning path →

03 / DIGITAL SOVEREIGNTY · OPEN TOOLS

What ZELYRA stands for:
a self-determined digital future.

ZELYRA stands for digital self-determination: open tools, readable source and local development. Our goal is to help people understand, shape and advance the software they rely on. The current source version is 1.2.40; supported workflows and checks are described in the documentation.

Zero-Tolerance for Flaws

The compiler checks implemented language constructs and reports detected issues. This is not a formal proof that a program is correct.

Excellence & Sovereignty

The compiler and Studio run locally. Review release notes for supported environments and database workflows.

Logical Transparency

Types and schema declarations make some expectations visible in source. The available checks depend on the construct.

Yield & Productivity

Studio can inspect supported schemas and preview generated CRUD source before you apply it. It does not change database schemas or records.

Resilience (Forged in Rust)

Database backends and web features have separate support boundaries. Check compatibility and deployment requirements for your use case.

Autonomy (Data Sovereignty)

Project imports link local source files. Bundle previews use a statically recognized subset and may omit dynamic dependencies or runtime configuration.

Reviewable by design: Inspect and edit generated source before applying it to your project.

🛡️ Local project tooling

⚡ Project status

Development note: Stable releases are available. This status does not certify an individual application or deployment; check the published feature limits and test the workflows your project depends on.

04 / getting started

The best way to learn
is by doing.

Four accessible steps guide you from your first program to a database-backed web application.

your first lesson

Say what you mean.

Zelyra code describes clearly what values a function expects and what result it delivers.

View example →

fibonacci.zyl

fn fibonacci(n: Int) -> Int {
    if n <= 1 {
        return n
    }

    return fibonacci(n - 1) +
           fibonacci(n - 2)
}

→ 55

05 / try it out

Read. Modify.
Understand.

Here is a small Zelyra page. Change the code in the editor or the parameter in the route bar and watch how it turns into a safe HTML response in real time.

Quick test:

✦

Good to know & interactive This browser demo illustrates selected examples; it is not the Zelyra compiler or a full application runtime.

hello_web.zyl ✏️ editable ▶ Open in Playground

page "/hello/{name}" {
    html {
        <h1>Hello, {name}!</h1>
        <p>Served safely via Zelyra built-in HTTP engine.</p>
    }
}

GET http://localhost:3000/hello/

zelyra serve :3000

HTTP/1.1 200 OK ⚡ rendered in 0.8ms (auto-escaped)

06 / language comparison

Where Zelyra fits:
Safety without complexity.

Full side-by-side comparison →

vs. TypeScript Schemas and types in one language

Zelyra lets you describe data structures and constraints in its own source files. The compiler checks the language rules implemented in the selected release.

vs. PHP & Python A focused language for web and database applications

Zelyra includes a compiler and project templates, but its checks do not rule out runtime errors. No comparative performance benchmarks are published.

vs. Rust Designed for Web & Database Applications

Zelyra’s compiler and runtime are written in Rust. Zelyra programs use Zelyra’s own type system and runtime; Rust’s ownership and borrowing checks do not automatically apply to Zelyra code.

07 / installation & quickstart

Ready in minutes.
Run locally or with Docker.

The CLI can run as a standalone binary. Database-backed applications still need configured services such as MariaDB.

Option 01 · Standalone

Native CLI Installation

Clones the repository and builds the CLI from source. Build time and application performance depend on your system and workload.

$ git clone https://github.com/sf1976/zelyra.git
$ cd zelyra && ./install.sh

Next command: zelyra check fibonacci.zyl

Option 02 · Full Stack

Docker Compose & MariaDB

Builds and starts the generated web application and MariaDB services using the project’s Compose file.

$ zelyra new my-app --mariadb
$ cd my-app && docker compose --env-file .env -f docker-compose.mariadb.yml up -d --build

Then initialize the schema: docker compose --env-file .env -f docker-compose.mariadb.yml exec web zelyra db setup main.zyl

open source & community

Build your next project
with true independence.

Explore the handbook, inspect the open-source compiler on GitHub or start with our 5-minute interactive tutorial.

GitHub Repository ↗ Read Handbook →

Business appsin one clear language.

Learn Zelyra on the go.Even on the train without Wi-Fi.

One language.One connected workflow.

Readable by people.Ready for AI.

One model.Connected steps.

Less boilerplate.More business logic.

What ZELYRA stands for:a self-determined digital future.

The best way to learnis by doing.

Say what you mean.

Read. Modify.Understand.

Where Zelyra fits:Safety without complexity.

Ready in minutes.Run locally or with Docker.

Build your next projectwith true independence.

Legal Notice