unwind

🥈Silver

Unwind analyzes and documents legacy codebases to enable AI-driven rebuilds. Operations teams use it to create machine-readable documentation and phased rebuild plans. It connects to existing codebases and integrates with AI agents for modernization.

900Updated 4mo ago

Intermediate30min to implementautomation

Saves ~480 min per use

Quick InstallView Source

git clone https://github.com/cliftonc/unwind.git

Works with:

Claude

Overview

About This Skill

Unwind analyzes legacy codebases across five phases—discovery, layer analysis, gap detection, gap completion, and synthesis—to generate complete machine-readable documentation and strategic rebuild plans. It extracts repository structure, identifies technology stacks, and documents each architectural layer (database, domain model, service, API, frontend) in dependency order, preserving external API contracts, business logic, and data integrity. Operations teams and engineers use Unwind to create work lists for AI-driven modernization, translating systems into new frameworks while maintaining behavioral compatibility. The skill outputs architecture diagrams, layer-by-layer documentation, and a comprehensive REBUILD-PLAN.md that serves as a blueprint for system reconstruction.

How to Use

Install via `/plugin install https://github.com/cliftonc/unwind` and restart Claude Code. Run `unwind:start` to discover architecture, `unwind:unwinding-codebase` to analyze layers, and `unwind:synthesizing-findings` to generate the final REBUILD-PLAN.md. Review intermediate outputs in `docs/unwind/` at each phase.

Use Cases

Documenting legacy applications before framework or language migrations

Creating AI-readable blueprints for system modernization projects

Generating layer analysis for monolith-to-microservices refactors

Building rebuild plans that preserve business logic and API contracts during technology transitions

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/cliftonc/unwind

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Analyze the following legacy codebase for [COMPANY] in the [INDUSTRY] sector. Generate machine-readable documentation and a phased rebuild plan. Focus on [DATA] as a critical component. Provide a summary of key findings and recommendations.

Example Output

# Legacy Codebase Analysis for GreenTech Solutions

## Key Findings
- **Codebase Age**: 15 years old, primarily written in COBOL and Fortran
- **Critical Systems**: 80% of core business logic resides in legacy systems
- **Dependencies**: 32 external integrations, 15 of which are deprecated
- **Security Risks**: 48 identified vulnerabilities, 12 critical

## Phased Rebuild Plan

### Phase 1: Documentation & Assessment (Weeks 1-4)
- Complete machine-readable documentation
- Identify critical paths and dependencies
- Assess security vulnerabilities

### Phase 2: Modular Rebuild (Weeks 5-12)
- Rebuild core modules in modern languages (Python, Java)
- Maintain legacy system interfaces
- Implement security patches

### Phase 3: System Migration (Weeks 13-20)
- Gradual migration to new systems
- Parallel run legacy and new systems
- Full cutover and decommission legacy systems

## Recommendations
- Prioritize rebuilding payment processing and customer data modules
- Allocate additional resources to security remediation
- Consider cloud migration for improved scalability