reversa-migrate

Orchestrates a six-agent pipeline that converts legacy system specs into migration-ready specifications.

Updated Jun 26, 2026
One-click install
npx skills add https://github.com/RLuf/fzagent --skill reversa-migrate-rluf
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: reversa-migrate
Source: https://github.com/RLuf/fzagent/tree/main/.agents/skills/reversa-migrate
Command: npx skills add https://github.com/RLuf/fzagent --skill reversa-migrate-rluf

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Migrating a legacy system to a modern stack requires coordinated decisions about paradigm, architecture, screens, and parity testing. This Skill orchestrates that entire migration pipeline after the Reversa discovery phase has populated _reversa_sdd/, so no decision step is skipped or done out of order. ## Core Features & Use Cases - Six-agent orchestration: Sequentially runs Paradigm Advisor, Curator, Strategist, Designer, Screen Translator, and Inspector with human decision gates between agents. - Resumable state machine: Tracks progress in .state.json with --resume, --regenerate=<agent>, and per-phase regeneration for Designer and Screen Translator. - Auto mode with logged defaults: --auto applies documented defaults without pauses and records every auto-decision in ambiguity_log.md for later review. - Use Case: After running /reversa on a legacy ERP, invoke /reversa-migrate to collect a migration brief, decide paradigm and topology, translate screens, generate parity tests, and produce a handoff.md ready for a coding agent. ## Quick Start Run /reversa-migrate after /reversa has populated _reversa_sdd/ to start the guided migration pipeline and produce the final handoff document.

Frequently Asked Questions about reversa-migrate

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I migrate a legacy system to a modern stack with AI agents?▼

Run /reversa first to generate legacy specs in _reversa_sdd/, then invoke /reversa-migrate. It collects a migration brief, runs six specialized agents in sequence with human approval gates, and produces a handoff.md for the coding agent.

How to resume an interrupted migration pipeline?▼

Use the --resume flag. The orchestrator reads _reversa_sdd/migration/.state.json, identifies the current agent, phase, and status, and continues from where it stopped, including re-running any pending human approval pause.

Can I regenerate only one agent's output in the migration pipeline?▼

Yes, use --regenerate=<agent> to redo an agent and all downstream agents, or --regenerate=designer:architecture to redo only a specific phase. A timestamped backup is created before any destructive operation.

What happens if the legacy system has no user interface?▼

The Screen Translator agent detects the absence of UI and completes with status skipped, producing no screen artifacts. The pipeline continues to the Inspector, and handoff.md notes that the system has no screens.

What does --auto mode do in a migration workflow?▼

Auto mode skips all human pauses and applies documented defaults, such as adopting the target stack's natural paradigm and the recommended strategy. Every auto-applied decision is logged in ambiguity_log.md for later human review.