Skip to content

mirako-studio/lumiko

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Lumiko

Auto-generate documentation from your codebase using Claude.

Lumiko scans your project, sends the source to Claude, and produces structured docs — README, architecture, API reference, and an AI-optimized context.json that agents can consume directly.

No custom parsers. No AST walking. Claude reads your code and writes the docs.

How it works

your codebase ──> Lumiko scans files ──> Claude analyzes ──> docs/ written

Lumiko uses Claude Code (your existing subscription) by default. No API key needed — just have the Claude Code CLI installed.

Quick start

npm install -g lumiko

cd your-project
lumiko init        # creates .lumiko/config.yaml
lumiko generate    # generates docs/

That's it. Your docs/ folder now has:

File What it is
README.md Project overview, install instructions, usage
architecture.md System design, Mermaid diagrams, module breakdown
api.md Exported functions, types, signatures, examples
context.json AI-optimized — every file mapped with purpose, exports, and function signatures

The context.json

This is what makes Lumiko different. It's not just human docs — it produces a structured JSON file designed for AI agents:

{
  "project": {
    "name": "your-project",
    "summary": "What this project does",
    "language": "TypeScript",
    "runtime": "Node.js 18+"
  },
  "architecture": {
    "pattern": "How the system is structured",
    "entryPoints": ["src/index.ts"],
    "dataFlow": "How data moves through the system"
  },
  "files": [
    {
      "path": "src/core/engine.ts",
      "purpose": "Main processing engine",
      "exports": ["Engine", "createEngine"],
      "keyFunctions": [
        { "name": "process", "signature": "process(input: Input): Output" }
      ]
    }
  ],
  "quickReference": {
    "commands": { "build": "npm run build", "test": "npm test" },
    "envVars": ["API_KEY", "DATABASE_URL"]
  }
}

Drop this into any agent's context and it instantly understands your codebase without reading every file.

CLI

lumiko init                        # set up config
lumiko init --force                # overwrite existing config

lumiko generate                    # generate docs (asks for confirmation)
lumiko generate -y                 # skip confirmation
lumiko generate --dry-run          # preview without calling Claude
lumiko generate --backend api      # use Anthropic API instead of Claude Code
lumiko generate --verbose          # show raw Claude output for debugging

Configuration

After lumiko init, edit .lumiko/config.yaml:

version: 1

project:
  name: my-project
  description: "Optional description"

# what to scan
include:
  - "src/**/*"
  - "lib/**/*"
  - "*.ts"

# what to skip
exclude:
  - "node_modules/**"
  - "dist/**"
  - "**/*.test.ts"

output:
  directory: docs
  formats:
    - markdown   # README, architecture, api
    - context    # context.json for AI agents

docs:
  readme: true
  architecture: true
  api: true
  diagrams: true

claude:
  backend: claude-code   # or "api" (requires ANTHROPIC_API_KEY)
  model: claude-sonnet-4-20250514
  maxTokens: 8192

Backends

Backend How it works What you need
claude-code (default) Runs claude --print under the hood Claude Code CLI installed
api Calls the Anthropic API directly ANTHROPIC_API_KEY env var

Claude Code uses your existing subscription — no extra API costs.

Requirements

  • Node.js 18+
  • Claude Code CLI (npm i -g @anthropic-ai/claude-code) or an Anthropic API key

License

MIT


Built by Mirako Studio

About

Documentation that updates itself. CLI tool for auto-generating docs from your codebase.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors