ct-writing-plans

Start with a header stating the goal, architecture, and tech stack, then decompose the spec into tasks with exact file paths and interfaces. Each task contains bite-sized TDD steps: write the failing test, verify it fails, implement minimally, verify it passes, and commit.

Each task should be the smallest unit carrying its own test cycle, with steps of 2-5 minutes each. Include exact file paths, complete code blocks, exact commands with expected output, and an Interfaces block describing what the task consumes and produces.

Split when the spec covers multiple independent subsystems. Each plan should produce working, testable software on its own, so one plan per subsystem keeps tasks focused and independently executable.

Plans fail when they contain placeholders like TODO or vague steps such as add error handling, when code is referenced but never shown, or when type and function names are inconsistent across tasks. A self-review pass for coverage, placeholders, and type consistency catches these.

After saving the plan, choose subagent-driven execution, which dispatches a fresh subagent per task with review between tasks, or inline execution, which runs tasks in the current session with batch checkpoints.