skills.rest
danielsuguimoto
danielsuguimotoCOMMUNITY
Explore all skills by this author →
𝕏@djsuguimotoGitHub

spawn

Runs a target skill or raw prompt inside an isolated subagent context.

Updated Jul 11, 2026
Search Tags:#agent orchestration#subagent#task delegation#skill invocation#context isolation#prompt execution
One-click install
npx skills add https://github.com/danielsuguimoto/skills --skill spawn-danielsuguimoto
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: spawn
Source: https://github.com/danielsuguimoto/skills/tree/main/skills/spawn
Command: npx skills add https://github.com/danielsuguimoto/skills --skill spawn-danielsuguimoto

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Long-running or context-heavy tasks pollute the parent agent's context window, making it harder to stay focused on the main conversation. This Skill offloads work to a subagent so only the final result returns to the parent. ## Core Features & Use Cases - Skill Invocation in a Subagent: Pass a skill name plus arguments and the subagent invokes that skill in its own context, keeping the skill body out of the parent. - Raw Prompt Execution: Pass any plain-language prompt and the subagent executes it directly with front-loaded context. - Automatic Mode Detection: The harness infers whether the input is a skill invocation or a raw prompt by checking the leading token against registered skill names. - Use Case: You want to run a verbose codebase audit without filling your main session with file contents. Spawn a subagent to run the audit skill and receive only its summary. ## Quick Start Ask the agent to spawn a subagent to run the debug skill on the failing login flow and return only the diagnosis.

Frequently Asked Questions about spawn

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

FAQPage Schema
How do I run a skill inside a subagent?▼

Pass the target skill's name followed by its arguments as the input. The harness matches the leading token against registered skill names and spawns a subagent that invokes that skill, returning only the result to the parent.

How does the spawn skill decide between skill invocation and raw prompt?▼

It inspects the input's leading token. If that token matches a known skill name via the harness's skill-listing mechanism, it treats the input as a skill invocation; otherwise the entire input is executed as a raw prompt.

Does the subagent share context with the parent agent?▼

No. Subagents are stateless and their working context never enters the parent. You must front-load all needed context into the subagent's prompt, and the parent receives only the subagent's final return.

What subagent profile does the spawn skill use by default?▼

It defaults to a general-purpose profile with full tool access running in the foreground, discovered through the harness's native subagent mechanism. No explicit configuration is required.

When should I not use subagent spawning for a task?▼

Avoid it for quick, low-context tasks where spawning overhead exceeds the benefit, or when the task needs continuous back-and-forth with the parent conversation, since subagents are stateless and return only a single result.

Related Skills

Explore More Software Engineering Skills→
CodeAlive-AICodeAlive-AI
119

subagents-management

Manage subagent definitions and skill packages across Claude Code, Codex CLI, and OpenCode.

Official
Intermediate
am-willam-will
1.0k

codex-subagent

Spawn Codex subagents for parallel research and multi-step workflows.

Community
Advanced
Liuchun-ossLiuchun-oss
15

dispatching-parallel-agents

Coordinate multiple specialized AI agents on independent problems without shared state.

Community
Intermediate
harsh-pandheharsh-pandhe

using-superpowers

Invokes the appropriate Skill tool before every conversational response.

Community
Advanced