vetcoders-spawn

Spawn external sub-agents in interactive zsh shells with durable artifacts.

1|Updated Mar 8, 2026
One-click install
npx skills add https://github.com/VetCoders/vetcoders-skills --skill vetcoders-spawn
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: vetcoders-spawn
Source: https://github.com/VetCoders/vetcoders-skills/tree/main/vetcoders-spawn
Command: npx skills add https://github.com/VetCoders/vetcoders-skills --skill vetcoders-spawn

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill allows you to delegate tasks to external AI agents, running them in isolated processes with visible or headless terminals, and collecting their durable artifacts for review.

Core Features & Use Cases

  • Parallel Task Delegation: Distribute work across multiple agents for faster completion.
  • Process Isolation: Run agents in separate shells, ensuring a clean environment.
  • Durable Artifacts: Generate reports, transcripts, and metadata for auditability.
  • Use Case: You need to research three different market segments for a new product. Use this Skill to spawn three agents, each tasked with researching one segment, and then synthesize their findings.

Quick Start

Use the codex_spawn.sh script to implement the plan found in '.ai-agents/plans/market-research.md' in a terminal window.

Frequently Asked Questions about vetcoders-spawn

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

FAQPage Schema
How do I delegate parallel tasks to external AI agents in isolated shells?

You can delegate parallel tasks by spawning external sub-agents through portable scripts that leverage an explicit spawn contract over zsh -ic, providing each agent a fresh interactive shell for isolated execution.

Can I run spawned agents in headless mode instead of a visible terminal?

Yes, spawned agents support both visible Terminal.app and headless modes, allowing you to choose whether to observe the local CLI runtime visually or run tasks silently in the background.

How does process isolation work when spawning multiple agents for parallel work?

Process isolation works by giving each spawned agent a fresh interactive shell via zsh -ic, ensuring a clean environment where agents observe the local CLI as a user would without interfering with each other.

Where do spawned agents write their reports and transcripts?

Spawned agents write durable report, transcript, and meta artifacts into the .ai-agents/ directory, providing complete auditability and reviewable outputs for delegated exploration and implementation tasks.

Do I need zsh to use the spawn contract for external agent delegation?

Yes, the spawn contract explicitly leverages zsh -ic to provide each external sub-agent with a fresh interactive shell, making zsh a required environment for executing the portable scripts.

What is the best way to automate research across multiple market segments using agents?

The best way is spawning multiple agents via the codex_spawn.sh script, assigning each agent one segment to research, then synthesizing the durable artifacts they generate into consolidated findings.