subagent-driven-development

Dispatch isolated subagents per task with two-stage spec and code-quality reviews.

60|38|Updated Feb 9, 2026
One-click install
npx skills add https://github.com/abcnuts/manus-skills --skill subagent-driven-development-abcnuts
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/abcnuts/manus-skills/tree/main/skills/workflow-skills/subagent-driven-development
Command: npx skills add https://github.com/abcnuts/manus-skills --skill subagent-driven-development-abcnuts

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Orchestrate complex task execution by coordinating multiple, isolated subagents per task, ensuring quality through two-stage reviews.

Core Features & Use Cases

  • Per-task subagent orchestration with isolation
  • Two-stage review workflow (spec compliance, code quality)
  • Fast iteration with parallelizable tasks
  • Self-contained prompt templates and example workflows

Quick Start

Initialize a plan, dispatch an implementer subagent for each task, then route through spec and code-quality reviewers until all tasks are complete.

Frequently Asked Questions about subagent-driven-development

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I orchestrate AI agents for task execution with code review?

Orchestrate AI agents by dispatching a fresh subagent per task, then routing the output through two-stage reviews for spec compliance and code quality.

What is two-stage code review in AI agent workflows?

Two-stage code review routes subagent output through a spec reviewer and a code-quality reviewer to ensure verifiable, high-quality implementations.

How do I isolate context for parallel tasks in AI automation?

Achieve context isolation by dispatching fresh subagents per task, ensuring each implementation runs independently with isolated context.

Does subagent-driven development work for executing plans in a single session?

Yes, it is designed for executing plans within a single session, utilizing per-task context isolation and an iterative approval loop.

Can I automate spec compliance checks for AI-generated code?

Yes, a dedicated spec reviewer subagent verifies implementations against requirements before final approval in the two-stage review workflow.