subagent

Name: subagent
Availability: InStock
Author: 42ATOM

Delegate complex tasks to codex or claude-code subagents with structured subtask cards.

Updated Jan 9, 2026

Search Tags:#workflow #monitoring #delegation #subagent #ai-assistant #task-management

Ready to run

npx skills add https://github.com/42atom/msgcode --skill subagent-42atom

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: subagent
Source: https://github.com/42atom/msgcode/tree/main/src/skills/optional/subagent
Command: npx skills add https://github.com/42atom/msgcode --skill subagent-42atom

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

主脑需要将复杂任务拆分并交给子代理执行，以减轻认知负担并确保结果可控与可追踪。

Core Features & Use Cases

教主脑如何把复杂任务拆给 codex / claude-code 这类执行臂，并在完成后进行验收。
适用于需要多阶段执行、审查、验收和结果回收的场景，例如代码评审、研究调查、文档审阅等。

Quick Start

Delegate a complex task to a codex or claude-code subagent and monitor its progress until completion.

Frequently Asked Questions about subagent

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

FAQPage Schema

How do I delegate complex tasks to subagents like codex or claude-code?▼

Task delegation to subagents uses a formal subtask card structure with unique tokens and acceptance criteria to monitor execution arms like codex or claude-code, ensuring verifiable results and reducing the main brain's cognitive workload.

What is the best way to monitor subagent task execution and verify results?▼

The best way to monitor subagent task execution is to enforce acceptance criteria and unique tokens for each delegated subtask. This ensures that multi-step tasks yield verifiable and controllable results upon completion.

When do I need to use subagent delegation for task management?▼

You need subagent delegation when managing multi-step tasks that require external execution arms, such as code评审, research investigations, or document reviews. It helps split complex work to ensure controllable and trackable outcomes.

Can I use this subagent workflow for code review and research investigation?▼

Yes, the subagent workflow is designed for scenarios requiring multi-stage execution, review, and acceptance, making it suitable for code reviews, research investigations, and document reviews to ensure formal result recovery.

Does delegating tasks to external execution arms reduce the main brain's workload?▼

Delegating tasks to external execution arms reduces the main brain's workload by splitting complex tasks into subtasks. The formal subtask card structure ensures these delegated tasks remain controllable and trackable.