Writing Implementation Tasks as Commands

Create implementation task breakdowns as Claude Code custom slash commands. Each task becomes an invokable /command that guides the agent through a single atomic implementation step.

Use this skill when breaking a feature spec and technical design into implementable tasks, planning implementation order, or defining what to build next.

Supporting files: CONTEXT-LAYERS.md for layer mapping details, EXAMPLES.md for complete examples.

Why Commands?

Implementation tasks as Claude Code commands provide:

• On-demand execution: User runs /task-1-data-model to start a specific task
• Self-contained context: Each command has all info needed for that task
• Clear workflow: Progress tracked via a overview command
• No wasted tokens: Only the current task is loaded, not the full task list

Output Structure