synthesis-code-planning

Code planning requires analyzing task descriptions and existing codebases to generate multiple implementation approaches with explicit pros and cons. This structured approach evaluates trade-offs and selects an optimal solution before generating any implementation code.

Evaluating implementation approaches involves generating at least two distinct strategies and comparing their explicit pros and cons. This forces analysis of architecture and maintainability trade-offs, allowing you to select the optimal solution with justification based on project constraints.

Introducing a new feature in a legacy codebase safely requires structured code planning to analyze existing conventions, propose multiple implementation strategies, and select the one with minimal regression risk. Implementation then follows existing conventions while adding error handling and tests.

Integrating a third-party library while weighing trade-offs requires generating multiple implementation approaches and evaluating their pros and cons against your architecture. You then select the optimal approach with justification and implement changes while marking modifications and following existing conventions.

Yes, structured code planning works for refactoring with minimal regression risk by forcing multi-approach evaluation and justified selection before implementation. It guides modifications by following existing conventions and adding error handling and tests to prevent regressions.