refactorMethodBody method

Future<AiGuidance> refactorMethodBody({ required String className, required List<String> stateFields, required String methodName, required String methodBody, NotifierType notifierType = NotifierType.stateNotifier, })

Returns AiGuidance for refactoring a single method body.

Sends a prompt to the local LLM; falls back deterministically when the LLM is unavailable.

Implementation

Future<AiGuidance> refactorMethodBody({
  required String className,
  required List<String> stateFields,
  required String methodName,
  required String methodBody,
  NotifierType notifierType = NotifierType.stateNotifier,
}) async {
  final prompt = _buildMethodPrompt(
    className: className,
    stateFields: stateFields,
    methodName: methodName,
    methodBody: methodBody,
    notifierType: notifierType,
  );
  final fallback = _fallbackMethodGuidance(
    className: className,
    stateFields: stateFields,
    methodName: methodName,
    notifierType: notifierType,
  );

  return _completeGuidance(
    title: 'Refactor $className.$methodName',
    subject: '$className.$methodName',
    category: 'logic-refactor',
    rationale:
        '$className.$methodName mutates local state and should move to immutable Riverpod state transitions.',
    prompt: prompt,
    fallbackRecommendation: fallback,
  );
}