skills.rest
Zpankz
ZpankzCOMMUNITY
Explore all skills by this author →
GitHub

subagent-driven-development

Dispatch fresh subagents per task with code reviews between tasks.

7|3|Updated Jan 15, 2026
Search Tags:#automation#orchestration#subagent#code-review#parallel-execution#development-workflow#task-dispatch
One-click install
npx skills add https://github.com/Zpankz/mcp-skillset --skill subagent-driven-development-zpankz
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: subagent-driven-development
Source: https://github.com/Zpankz/mcp-skillset/tree/main/subagent-driven-development
Command: npx skills add https://github.com/Zpankz/mcp-skillset --skill subagent-driven-development-zpankz

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill orchestrates plan execution by spawning fresh subagents for each task, enabling independent work in the same session and preventing shared-state contamination.

Core Features & Use Cases

  • Fresh subagent per task to avoid context pollution and enable rapid iteration.
  • Supports sequential, parallel, and parallel-investigation execution modes for diverse work scopes.
  • Built-in code reviews between tasks to catch issues early and maintain quality gates.
  • References and guidance for process-heavy workflows (sequential, parallel, and parallel investigation) to inform decision-making.

Quick Start

Provide a task plan and let the system dispatch fresh subagents with reviews for each task.

Frequently Asked Questions about subagent-driven-development

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

FAQPage Schema
How do I prevent context pollution when automating multiple development tasks in a single session?▼

To prevent context pollution during task automation, you can dispatch a fresh subagent for each task. This approach isolates independent work and stops shared-state contamination, enabling rapid iteration across diverse issues.

What is the best way to orchestrate parallel execution of independent coding tasks?▼

The best way to orchestrate parallel execution is using a subagent-driven workflow that supports parallel and parallel-investigation modes. This dispatches fresh subagents for concurrent task processing to accelerate development cycles.

How do code reviews fit into an automated task dispatch workflow?▼

Code reviews fit into automated task dispatch by acting as quality gates between tasks. The system spawns a fresh subagent per task and runs built-in code reviews sequentially to catch issues early and maintain quality.

Can I run sequential and parallel investigation modes for process-heavy development workflows?▼

Yes, you can run sequential, parallel, and parallel-investigation execution modes. The system provides specific references and guidance to help you manage process-heavy workflows and inform your execution mode decisions.

Why should I use fresh subagents for each task instead of a single agent for an entire plan?▼

You should use fresh subagents for each task to avoid context pollution and prevent shared-state contamination. This enables independent work within the same session and maintains high-quality results through built-in review gates.

Related Skills

Explore More Software Engineering Skills→
m16khbm16khb

subagent-driven-development

Dispatch fresh subagents per task with code review and test gates.

Community
Advanced
jenswittmannjenswittmann
10

build

Automate project builds and asset compilation in CI/CD pipelines.

Community
Basic
pacaplanpacaplan
30

implement-task

Dispatch autonomous subagents for sequential task implementation with TDD enforcement.

Community
Advanced
jg-chalk-iojg-chalk-io

moai-cc-hook-model-strategy

Automate Claude Code hook enforcement for validation, sequencing, and orchestration.

Community
Advanced