subagent-driven-development

Dispatch isolated subagents per task with automated reviews and progress tracking.

26|1|Updated Mar 19, 2026
One-click install
npx skills add https://github.com/BinyaminEden/claude-combine --skill subagent-driven-development-binyamineden
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/BinyaminEden/claude-combine/tree/main/skills/subagent-driven-development
Command: npx skills add https://github.com/BinyaminEden/claude-combine --skill subagent-driven-development-binyamineden

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Manually executing implementation plans is slow, context-heavy, and prone to missed requirements or inconsistent code quality, especially for plans with multiple independent tasks. This skill eliminates that overhead by automating task dispatch, review, and progress tracking, so you can focus on coordination instead of implementation details.

Core Features & Use Cases

  • Per-task subagent dispatch: Assigns a fresh, focused subagent to each independent task with exactly the context it needs, avoiding context pollution from prior tasks or session history.
  • Automated review gates: Runs spec compliance and code quality reviews after every task, plus a final whole-branch review before merge, catching gaps and issues early.
  • Progress persistence: Tracks completed tasks in a durable ledger that survives context compaction, so you never redo work after a session reset.
  • Use Case: For a 10-task feature implementation plan with decoupled tasks, this skill executes all tasks in a single session, automatically reviews each for spec and quality, and produces a merge-ready branch with no manual intervention between tasks.

Quick Start

Use the subagent-driven-development skill to execute the implementation plan stored at docs/feature-plan.md.

Frequently Asked Questions about subagent-driven-development

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

FAQPage Schema
How do I automate implementation plan execution for multiple independent tasks?

Automating implementation plan execution is done by dispatching isolated subagents per task, providing each with focused context to avoid pollution, and running automated spec compliance and code quality reviews before merge.

What is subagent-driven development and how does it handle context preservation?

Subagent-driven development assigns a fresh subagent to each task to prevent context compaction issues, using a durable progress ledger to track completed tasks so work is never redone after a session reset.

How do I run automated code review gates for every task in a feature plan?

Automated code review gates are run after every task for spec compliance and quality, followed by a final whole-branch review before merge, catching gaps and issues early without manual intervention.

Can I execute a 10-task feature implementation plan in a single session?

Yes, for a 10-task feature implementation plan with decoupled tasks, this approach executes all tasks in a single session, automatically reviewing each for spec and quality, and produces a merge-ready branch.

Does subagent-driven development support TDD and cost-optimized model selection?

Yes, this approach supports TDD workflows and cost-optimized model selection per task role, ensuring isolated subagents apply tests and reviews efficiently while preserving context across compaction.

What are the limitations of using isolated subagents for task automation?

This approach is limited to software development workflows with decoupled implementation tasks; tightly coupled tasks requiring shared context between steps may not suit isolated per-task subagent dispatch.