| name | extract-engine |
| type | atomic |
| license | MIT |
| description | Extracts existing Rails app code into a reusable engine incrementally — scaffolds engine structure, moves stable domain logic first, creates adapter interfaces to decouple host dependencies, and preserves regression coverage throughout each extraction slice. Each slice has one coherent responsibility, minimal new public API, passing regression tests, and a clear next step. Use when a developer needs to extract a feature into a Rails engine, move code out of a host app, decouple host coupling via adapters, or perform incremental extraction while preserving existing behavior. Trigger words: extract to engine, move feature to engine, host coupling, adapters, extraction slices, preserve behavior, incremental extraction.
|
| metadata | {"version":"1.0.0","user-invocable":"true"} |
Extract Engine
Use this skill when the task is to move existing code out of a Rails app and into an engine.
Prefer incremental extraction over big-bang rewrites. Preserve behavior first, then improve design.
Quick Reference
| Phase | Focus |
|---|
| Prep | Identify bounded feature and host dependencies |
| Logic | Move stable domain logic (POROs, services) first |
| Seams | Add adapters/config for host dependencies |
| Web | Move controllers, routes, views last |
HARD-GATE
DO NOT extract and change behavior in the same step.
Extraction must preserve existing behavior; refactoring and improvements belong in a separate step after the move is complete and verified.
Core Process
- Identify the bounded feature to extract — one coherent responsibility.
- List hard dependencies on the host app (models, services, config).
- Define the future engine boundary and host contract.
- Move stable domain logic first: POROs, services, value objects, policies, query objects. Delay direct host model references, authentication, route ownership, and asset integration.
- Add adapters or configuration seams for host-owned dependencies. Replace hardcoded host references with config values, adapter objects, or service interfaces.
- Move controllers, routes, views, or jobs only after seams are clear.
- Keep regression coverage green throughout each slice.
Each slice must have: one coherent responsibility, minimal new public API, passing regression tests, and a clear next step.
Extended Resources
Pitfalls
| Pitfall | What to do |
|---|
| Extracting too much at once | One bounded slice per step |
| Direct host references in engine | Use adapters or config |
| Behavior changes mixed with extraction | Preserve behavior first; refactor after |
| Circular dependencies introduced | Verify import graph before each slice |
| Implicit host contract | Explicitly document and test the host app contract |
Examples
First slice (move PORO, no host model yet):
mkdir -p my_engine/app/services/my_engine
mv app/services/pricing/calculator.rb my_engine/app/services/my_engine/pricing_calculator.rb
module MyEngine
class PricingCalculator
def initialize(line_items)
@line_items = line_items
end
def total
@line_items.sum { |item| item.price * item.quantity }
end
end
end
Verify regression coverage still passes before proceeding:
bundle exec rspec spec/services/pricing/ spec/requests/orders/
Adapter for host dependency:
module MyEngine
def self.current_user_for(request)
config.current_user_provider.call(request)
end
end
OrderCreator.for_request(request)
Output Style
- Propose one small, bounded extraction slice at a time.
- Outline the files moving, the new boundaries, and the regression tests to run.
- Language — Must be in English unless explicitly requested otherwise.
Integration
| Skill | When to chain |
|---|
| create-engine | Engine structure, host contract, namespace design after extraction |
| test-engine | Dummy app, regression tests, integration verification |
| refactor-code | Behavior-preserving refactors before or after extraction slices |