Apogee

The programming language built for the AI era. Compile-time safety. Intent verification. Runs everywhere.

@intent("greet the user by name, never null")
fn greet(name: String) -> String {
  "Hello, \(name)! Welcome to Apogee."
}

type User {
  name: String
  age: Int where age >= 0
}

let user = User { name: "Tyler", age: 35 }
print(greet(user.name))

$ apogee run hello.apg
Hello, Tyler! Welcome to Apogee.

---

Install

pip install apogee-lang

Or from source:

git clone https://github.com/apogee-lang/apogee.git
cd apogee
pip install -e .
apogee --help

CLI

Command	Description
apogee compile <file.apg>	Transpile to Python (.py)
apogee run <file.apg>	Compile and execute immediately
apogee check <file.apg>	Type-check only, no output

Why Apogee?

1. Null safety at compile time

let name: String = get_name()    // must return String, not null
let maybe: String? = find_user() // nullable — callers must handle it
print(maybe?.name)               // safe access with ?

2. Constraint types

type User {
  name: String
  age: Int where age >= 0    // compiler rejects User { age: -1 }
}

The compiler catches User { age: -1 } at build time. Runtime values are validated at construction.

3. @intent annotations

@intent("sort users by age descending, preserve original list")
fn sorted_by_age(users: [User]) -> [User] { ... }

Machine-readable contracts that AI tools can verify against the implementation.

4. Structured concurrency

spawn {
  fetch_users()
  fetch_posts()
}
// Both complete before execution continues. No dangling tasks.

5. Query expressions

let adults = from users where it.age >= 18

Read more: Why Apogee? | Comparison table vs Python, TypeScript, Rust, Go

Documentation

Document	Description
Language Spec	Full EBNF grammar, type system rules, semantics
Why Apogee?	Design rationale and language comparison
Contributing	Build from source, add features, code style
Changelog	Release history
Roadmap	Python → LLVM → WASM → JVM

Running Tests

python -m tests.test_runner

20 test programs covering all language features: functions, types, constraints, string interpolation, query expressions, spawn blocks, @intent, and 5 compile-time error tests.

Contributing

See CONTRIBUTING.md. The short version:

1. Fork and clone
2. pip install -e .
3. Make changes following the pipeline: spec → lexer → parser → typechecker → emitter → tests
4. python -m tests.test_runner — all green
5. Open a PR

Issues labeled good first issue are a great place to start.

Roadmap

Phase	Target	Timeline
1	Python transpiler	Now
2	Native LLVM backend	Month 6
3	WebAssembly target	Month 12
4	JVM target	Month 18

See ROADMAP.md for details.

License

MIT