subagents

Spawn isolated subagents for parallel or long-running background tasks.

14|6|Updated Feb 13, 2026
One-click install
npx skills add https://github.com/jholhewres/devclaw --skill subagents-jholhewres
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: subagents
Source: https://github.com/jholhewres/devclaw/tree/main/pkg/devclaw/copilot/builtin/skills/subagents
Command: npx skills add https://github.com/jholhewres/devclaw --skill subagents-jholhewres

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Enables running parallel, long-running, or complex tasks in isolated worker agents so the main session remains responsive and uncluttered, preventing task overload and context leakage.

Core Features & Use Cases

  • Spawn and manage isolated workers: Create subagents for background research, skill creation, or long-running analysis without blocking the primary agent.
  • Inter-agent communication: Discover sessions, send messages, and export conversation histories to coordinate work across agents or channels.
  • Monitoring and control: List running subagents, wait for results with timeouts, and stop stuck agents to reclaim resources.

Quick Start

Spawn a subagent to research GraphQL pagination practices and wait for its result.

Frequently Asked Questions about subagents

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

FAQPage Schema
How do I run parallel tasks in background processing without blocking the main agent session?

To run parallel tasks without blocking the main session, you spawn isolated subagents that execute long-running operations independently. This background processing approach prevents task overload and context leakage while keeping your primary session responsive.

What is multi-agent orchestration and how do isolated workers coordinate sessions?

Multi-agent orchestration coordinates isolated worker agents across separate sessions to handle complex tasks. You can discover active sessions, send inter-agent messages, and export conversation histories to synchronize work and share results across agents.

How do I monitor background subagents and reclaim resources from stuck parallel tasks?

To monitor background subagents, you list all running instances, wait for results with configurable timeouts, and stop stuck agents to reclaim resources. This control mechanism ensures long-running parallel tasks do not block your workflow indefinitely.

Can I use cross-agent messaging to coordinate parallel research tasks across multiple sessions?

Cross-agent messaging allows you to send messages between isolated subagents running parallel research tasks. By leveraging session management actions like list and send, you can coordinate multiple agents working on different research components simultaneously.

What are the limitations of spawning isolated agents for long-running analysis tasks?

Spawning isolated agents for long-running analysis requires active monitoring to avoid resource exhaustion. You must manually manage timeouts and stop stuck subagents, as unmonitored background tasks can accumulate and consume available session resources.