subagent-driven-development

name	subagent-driven-development
description	Use when executing implementation plans with independent tasks in the current session

Subagent-Driven Development

Execute plan by dispatching fresh subagent per task, with two-stage review after each: spec compliance review first, then code quality review.

Core principle: Fresh subagent per task + two-stage review (spec then quality) = high quality, fast iteration

When to Use

• Have implementation plan
• Tasks mostly independent
• Want to stay in current session

The Process

1. Read plan, extract all tasks with full text, note context, create TodoWrite.
2. For each task:
   • Dispatch implementer subagent with full task text + context
   • Answer subagent questions if any
   • Let implementer implement, test, self-review
   • Dispatch spec-review pass
   • If spec issues found: resume implementer, then re-review
   • Dispatch code-quality review pass
   • If quality issues found: resume implementer, then re-review
   • Mark task complete in TodoWrite
3. After all tasks, dispatch final code-reviewer for entire implementation.
4. Use finishing-a-development-branch.

Advantages

• Fresh context per task
• Review checkpoints automatic
• Spec compliance prevents over/under-building
• Code quality review improves maintainability

Red Flags

Never:

• Skip reviews
• Proceed with unfixed issues
• Dispatch multiple implementation subagents in parallel if they may conflict
• Move to next task while either review has open issues

Integration

Required workflow skills:

• writing-plans
• requesting-code-review
• finishing-a-development-branch

Subagents should use:

• test-driven-development