Limo CLI

Command-line interface for Limo - AI-powered codebase analysis and migration planning tool.

Features

• 🤖 Multi-Agent Analysis - Planner, Analyst, Writer, and Editor agents collaborate
• 📊 Professional Reports - Executive summaries, diagrams, and detailed analysis
• 🎯 LIMO.md Support - Customize analysis scope with LIMO.md configuration
• 📈 Semantic Diagrams - Type-safe architecture diagrams with nodes and edges
• 🌍 Multi-language - Supports English, 简体中文, 日本語, and more

Installation

npm install
npm run build

Usage

Basic Analysis

node dist/index.js analyze /path/to/project

With Options

# Specify language
node dist/index.js analyze /path/to/project --lang 简体中文

# Verbose mode (show AI thinking)
node dist/index.js analyze /path/to/project --verbose

# Custom modules
node dist/index.js analyze /path/to/project --modules architecture,security,dependencies

# Custom output directory
node dist/index.js analyze /path/to/project --output .limo-reports

LIMO.md Configuration

Create a LIMO.md file in your project root to customize the analysis:

# LIMO Configuration

## Focus Areas
- security
- architecture
- dependencies

## Exclude
- testing
- migration

## Analysis Scope
Focus on the authentication and authorization modules.

## Company Context
This is a financial institution with strict security requirements.

## Constraints
- Must use Java 17
- Cannot use external dependencies
- Must maintain backward compatibility

The CLI will automatically detect and parse LIMO.md, adjusting the analysis accordingly.

Architecture

4-Phase Analysis Workflow

1. Planning - Planner creates comprehensive analysis plan
2. Analysis - Analyst performs deep code exploration
3. Writing - Writer generates detailed report sections
4. Editing - Editor creates executive summary and finalizes report

Report Structure

# Analysis Report - Project Name

**Generated**: February 21, 2026, 12:00:00 AM
**Project**: /path/to/project
**AI Model**: claude-opus-4.6
**Analysis Modules**: architecture, security, dependencies

**Statistics**:
- Report Sections: 15
- Diagrams: 8
- Key Findings: 42

---

## 📊 Executive Summary

[Overall analysis summary...]

---

## Architecture Analysis

[Detailed content...]

![Architecture Diagram](data:image/svg+xml;base64,...)

*Diagram: System architecture showing main components*

---

## Security Analysis

[Detailed content...]

![Security Flow](data:image/svg+xml;base64,...)

*Diagram: Authentication and authorization flow*

Report Output

Reports are saved to .limo/reports/{uuid}/:

• report.md - Final markdown report
• memory.json - Analysis memories
• plan.json - Analysis plan

Development

Build

npm run build

Watch Mode

npm run watch

Debug

# Enable debug logging
LIMO_DEBUG=true node dist/index.js analyze /path/to/project

# Enable API debug logging
LIMO_DEBUG_API=true node dist/index.js analyze /path/to/project

Requirements

• Node.js 18+
• GitHub Copilot subscription (for AI analysis)
• gh CLI installed and authenticated

Project Structure

src/
├── index.ts              # CLI entry point
├── commands/
│   └── analyze.ts        # Main analysis command
├── agents/
│   ├── planner.ts        # Planning agent
│   ├── analyst.ts        # Analysis agent
│   ├── writer.ts         # Writing agent
│   └── editor.ts         # Editing agent
├── tools/
│   ├── index.ts          # Core tools
│   ├── extended.ts       # Extended tools
│   └── additional.ts     # Additional tools
├── report/
│   ├── markdownGenerator.ts    # Report generation
│   ├── diagrams.ts             # Diagram embedding
│   └── graphCompiler.ts        # Graph to SVG compilation
├── utils/
│   ├── debug.ts                # Debug utilities
│   └── limoConfigParser.ts     # LIMO.md parser
└── types/
    └── graphSemantics.ts       # Graph type definitions

prompts/
├── planner.md            # Planner agent prompt
├── analyst.md            # Analyst agent prompt
├── writer.md             # Writer agent prompt
└── editor.md             # Editor agent prompt

Key Features

Semantic Diagrams

Uses type-safe graph semantics for diagrams:

{
  "diagram_id": "architecture",
  "title": "System Architecture",
  "nodes": [
    { "id": "web", "label": "Web Layer", "type": "layer" },
    { "id": "api", "label": "API Layer", "type": "layer" },
    { "id": "data", "label": "Data Layer", "type": "layer" }
  ],
  "edges": [
    { "from": "web", "to": "api", "type": "calls" },
    { "from": "api", "to": "data", "type": "calls" }
  ]
}

Available node types: component, layer, system, data, process, actor, external Available edge types: dependency, calls, uses, contains, inherits, implements, data_flow, control_flow, stored_in, manages

LIMO.md Parser

Automatically parses LIMO.md with smart detection:

• Structured Markdown - Sections with headings and lists
• Natural Language - Keywords like "focus on", "skip", "must"
• Module Validation - Validates against known modules
• Constraints Extraction - Finds "must", "cannot", "should not" statements

License

MIT

Contributing

Contributions welcome! Please open an issue or PR.

Built with ❤️ using GitHub Copilot and Claude