subagent-driven-development

Installation
SKILL.md

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

Scope Boundary and Routing

  • Use this skill when a concrete implementation plan already exists and tasks should be executed in controlled sequence with review gates.
  • Do not use this as the first choice for scattered, unrelated test failures without a plan.
  • For parallel root-cause investigations across independent failures, route to dispatching-parallel-agents.

When to Use

Installs
2
First Seen
Mar 1, 2026
subagent-driven-development — tyk-lab/my-ai-skill