llm-as-code 0.1.4

LaC (LLM as Code) - reference engine. Declarative behavior for LLM apps: law in files, perimeter in code.

LaC - LLM as Code

LaC is a protocol for building LLM applications the way IaC builds infrastructure: behavior is declared in versioned artifacts, not improvised in prompts. An application is a folder of plain files - a compose declaration, law files with explicit authority levels, and a commands module. The engine is generic and knows nothing about any particular application.

This repository holds the LaC specification (SPEC.md) and the reference engine. The engine is intentionally small: a compose loader with a strict parser, a context assembler that loads law in trust order (L1 then L2 then L3), a REPL, and an agentic loop where free-form language and canonical !commands are two roads into the same code.

Status: pre-release. The spec (SPEC.md, methodology v0.2) is the canon; this engine tracks it through milestones M0-M3 (loader, retrieval commands, write perimeter, drift metrics). APIs and the compose schema will change.

Core ideas

• Levels are roles. L1 (admin) and L2 (dev) files carry behavior and load always; L3 (user data) is content, never instructions.
• Two roads, one code. Canonical !commands bypass the model entirely (zero tokens); free-form requests reach the same functions through tool calling. The model interprets and narrates; code executes.
• Perimeter in code, not in prose. Effects on disk are the engine's job. A wrong action must be survivable.

Quick start

pipx install llm-as-code
cd your-app
lac

An application folder looks like:

your-app/
  .lac/
    llm_compose.yaml    # the anchor: all paths resolve relative to the app root
    law/                # L1/L2 law files
    souls/              # personas (L2)
  commands_<app>.py     # the app's command module (TOOLS + COMMANDS registry)
  <memory folders>      # L3 content

The engine finds .lac/llm_compose.yaml in the current directory (or in the directory passed as the first argument) and resolves every path in the compose against that root. No absolute paths anywhere.

Providers: Anthropic (raw HTTP, prompt + conversation caching, ANTHROPIC_API_KEY from env only), Mistral (MISTRAL_API_KEY), and Ollama for local models. The vendor surface lives in one dispatch table, adapter.PROVIDERS, behind adapter.send(); any OpenAI-shaped API is one new entry.

Trust model: the commands module is application code (L2) and the engine executes it deliberately - running an app means trusting its dev layer, the same way installing any package means trusting its authors. The levels protect L1/L2 from the model and from L3 content, not the machine from the app you chose to run.

First application

Grimoire - a personal memory system and the first LaC application. Its earlier implementation runs the same protocol on Claude Code, with the perimeter enforced by harness permissions instead of engine code: one law, two hosts.

License

Apache-2.0. The protocol is free; applications built on it license themselves independently.