Skip to content

Source status: no status is stated in the source document, so none is assigned here. Do not treat this page as a finalized specification. The normative rules are in the Technical Specification.

Name: This language was renamed from AXIOM to KYLLO on 2026-11-08. The original source files in source-original/LANG/ still use the name AXIOM.

KYLLO is a programming language for LLM's

The purpose of this programming language is to make software development by AI agents more reliable, predictable, and verifiable.

The language is designed around the idea that AI-generated code should be checked against what it is supposed to accomplish, not merely whether it is syntactically valid.

  1. Prevent errors: Catch mistakes before they become runtime failures whenever possible.
  2. Enforce requirements: Allow developers to define rules that programs must obey.
  3. Verify behavior: Check whether functions and systems satisfy their declared requirements.
  4. Make failures explicit: Require programs to account for possible failure states rather than silently ignoring them.
  5. Control side effects: Make operations such as file access, database changes, and network requests explicit.
  6. Give AI useful feedback: Provide precise, structured compiler errors that AI agents can understand and correct.
  7. Prioritize correctness: Favor predictable, maintainable software over shorter code or faster code generation.

Core Philosophy

A program should not be considered correct simply because it compiles or runs.

It should be checked against its intended behavior, and anything that cannot be verified should be identified as unverified rather than assumed correct.

Primary Goal

Create a programming language in which AI agents can build software with fewer mistakes because correctness is built into the language itself, rather than left entirely to the agent's judgment.

Core Design Principles

  1. Correctness over convenience: The language should prioritize preventing incorrect software over making code shorter or easier to write.

  2. Explicit over implicit: Important behavior, permissions, dependencies, and failure conditions should be clearly declared.

  3. Verifiable by design: Requirements should be expressible in a form the compiler or verification tools can check.

  4. No assumed success: Anything that cannot be verified must be identified as unverified. The language must not silently treat uncertainty as success.

  5. Designed for AI agents: Syntax, errors, documentation, and compiler feedback should be structured so AI agents can reliably understand and work with them.

Source and authority

This page migrates MAIN.md in full. It is an explanatory / implementation document, not the normative specification. Where it simplifies a rule, the Technical Specification governs. The original file is preserved in source-original/LANG/.