writing-plans

To turn a spec into an implementation plan, provide the feature spec to generate a structured task breakdown with concrete file paths, testing steps, and explicit commit strategy. This ensures engineers can execute tasks with zero context by following bite-sized, reproducible steps saved as plan artifacts.

The best way to structure implementation plans for zero-context execution is using a fixed format containing a header, bite-sized task list, concrete file paths, testing steps, and explicit commit steps. This structure guarantees reproducible execution and clear ownership for engineers.

Yes, you can include testing steps and documentation mapping in your task breakdown. The generated plan enforces TDD-oriented planning by outlining necessary docs, stubs, and specific save locations like docs/plans for traceability, alongside acceptance criteria and verification steps.

To save implementation plans for traceability, generate the plan artifact and save it to the docs/plans directory using a date-prefixed naming convention like docs/plans/YYYY-MM-DD-<feature-name>.md. This ensures thorough documentation mapping and historical tracking of spec changes.

Yes, this planning approach works without requiring external dependencies. It operates independently to parse your feature spec and output a fixed plan structure with task lists and commit steps, requiring no additional components or environmental setup to generate the documentation artifacts.

You need to generate a fixed plan structure from a spec when your project requires clear, bite-sized task breakdowns and thorough documentation before any code changes. It is essential for guiding engineers with zero context and ensuring reproducible execution through explicit commit steps.