subagent-driven-development

name	subagent-driven-development
version	1.0.0
description	Use when executing implementation plans with independent tasks in the current session
dependencies	{"@skill-module-example/using-git-worktrees":"^1.0.0","@skill-module-example/writing-plans":"^1.0.0","@skill-module-example/requesting-code-review":"^1.0.0","@skill-module-example/finishing-a-development-branch":"^1.0.0","@skill-module-example/test-driven-development":"^1.0.0"}

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.

Why subagents: You delegate tasks to specialized agents with isolated context. By precisely crafting their instructions and context, you ensure they stay focused and succeed at their task.

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

When to Use

• Have an implementation plan with mostly independent tasks
• Want to stay in the current session
• Tasks can be dispatched to fresh subagents

vs. Executing Plans (parallel session):

• Same session (no context switch)
• Fresh subagent per task (no context pollution)
• Two-stage review after each task
• Faster iteration

The Process

1. Read plan, extract all tasks with full text, note context, create TodoWrite
2. For each task:
   • Dispatch implementer subagent
   • Handle any questions from implementer
   • Implementer implements, tests, commits, self-reviews
   • Dispatch spec reviewer subagent (confirm code matches spec)
   • Dispatch code quality reviewer subagent
   • Mark task complete
3. Dispatch final code reviewer for entire implementation
4. Use finishing-a-development-branch skill

Handling Implementer Status

• DONE: Proceed to spec compliance review
• DONE_WITH_CONCERNS: Read concerns before proceeding
• NEEDS_CONTEXT: Provide missing context and re-dispatch
• BLOCKED: Assess blocker and escalate appropriately

Model Selection

• Mechanical tasks (1-2 files, clear specs): use a fast, cheap model
• Integration tasks (multi-file coordination): use a standard model
• Architecture/design/review: use the most capable available model

Red Flags

Never:

• Skip reviews (spec compliance OR code quality)
• Proceed with unfixed issues
• Dispatch multiple implementation subagents in parallel
• Start code quality review before spec compliance is ✅
• Move to next task while either review has open issues