exo:spawn

Spawn parallel sub-agents with configurable depth and memory isolation.

1|Updated Feb 16, 2026
One-click install
npx skills add https://github.com/midsphere-ai/exo --skill exo-spawn
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: exo:spawn
Source: https://github.com/midsphere-ai/exo/tree/main/skills/spawn
Command: npx skills add https://github.com/midsphere-ai/exo --skill exo-spawn

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Orchestrates dynamic self-spawning of parallel sub-agents to tackle complex workloads while ensuring memory isolation and controlled execution depth.

Core Features & Use Cases

  • Self-spawn control: configure allow_self_spawn and max_spawn_depth to manage recursive spawning.
  • Parallel execution: spawn_self launches multiple child agents simultaneously.
  • Memory isolation & context fork: each child gets a fresh ShortTermMemory, while LongTermMemory remains shared; context can fork for isolation.
  • Safety guards: depth and per-call child limits prevent runaway spawning; provider-safe execution ensures safe multi-agent runs. Use Case: A coordinator delegates a set of subtasks to parallel children to accelerate processing and then synthesizes results.

Quick Start

Create a root coordinator with allow_self_spawn and spawn_self enabled, then call spawn_self with a list of tasks to run in parallel.

Frequently Asked Questions about exo:spawn

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

FAQPage Schema
How do I spawn parallel sub-agents to divide work and accelerate processing?

To spawn parallel sub-agents, call spawn_self with a list of tasks to launch multiple child agents simultaneously, accelerating processing and enabling result synthesis.

How does memory isolation work when spawning parallel child agents?

Memory isolation provides each spawned child a fresh ShortTermMemory while keeping LongTermMemory shared, ensuring isolated short-term context and accessible long-term knowledge across children.

Can I control the spawn depth to prevent runaway recursive agent spawning?

Yes, you can prevent runaway recursive spawning by configuring max_spawn_depth and per-call child limits as safety guards to strictly control execution depth and spawned agent count.

What's the best way to manage context forking when delegating subtasks to parallel agents?

The best way to manage context forking is to enable it for isolation, ensuring each parallel child agent operates within its own separated context while sharing long-term knowledge.

Does self-spawning support configurable limits for multi-agent runs?

Self-spawning supports configurable limits via allow_self_spawn and max_spawn_depth to manage recursive spawning and enforce execution boundaries, ensuring provider-safe multi-agent runs.