writing-plans

To generate an implementation plan from a software spec, provide the feature requirements to produce a detailed, executable document with exact file-level changes, test code, commands, and commit instructions. This breakdown eliminates ambiguous instructions and ensures reliable execution.

Breaking down feature development into atomic TDD steps involves structuring each task sequentially as write failing test, implement code, verify execution, and commit. This ensures precise file modifications and creates a self-check checklist to detect placeholders and mismatches with the spec.

Yes, you can use subagent execution workflows with implementation plans by supporting subagent-driven or inline execution options. The generated plans advise on using isolated worktrees and saving the documents under docs/superpowers/plans for reliable subagent implementation.

The best way to standardize engineering docs for task breakdown is to enforce a consistent plan document header containing the goal, architecture, tech stack, and file paths. This metadata provides essential context so any developer can execute the work without prior domain knowledge.

Yes, this implementation planning approach works for multi-component features by converting complex specs into a series of small, test-driven tasks. It specifies exact files to create or modify and maps out runnable test steps with expected outputs for consistent review.