spawn-agent

Spawn a Claude agent in a worktree for research or code review tasks.

5|1|Updated Dec 12, 2025
One-click install
npx skills add https://github.com/ncukondo/reference-manager --skill spawn-agent-ncukondo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: spawn-agent
Source: https://github.com/ncukondo/reference-manager/tree/main/.claude/skills/spawn-agent
Command: npx skills add https://github.com/ncukondo/reference-manager --skill spawn-agent-ncukondo

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Spawns a Claude agent in a worktree to handle research tasks, PR comments, or custom tasks.

Core Features & Use Cases

  • Spawns an agent into a dedicated worktree to perform research or code-review related tasks.
  • Supports PR-driven workflows, interactive prompts, and custom automation scenarios.
  • Provides flexible deployment across main repository or isolated worktrees to keep experiments isolated.

Quick Start

Launch a worker for a new feature by running the spawn-agent script with a feature branch and a task description.

Frequently Asked Questions about spawn-agent

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

FAQPage Schema
How do I automate code review and PR comments using a Claude agent?

You can automate PR comments by spawning a Claude agent in an isolated git worktree to handle code review tasks. The agent operates within a dedicated worktree to review changes and provide PR commentary without affecting your main repository.

Can I run repository research tasks in an isolated worktree without affecting the main branch?

Yes, spawning a Claude agent in an isolated worktree keeps research investigations separate from your main branch. This flexible deployment allows experiments and research tasks to run safely without modifying the primary repository state.

How do I spawn an agent for custom automation tasks across multiple projects?

Spawning an agent for custom tasks involves running a script with a feature branch and a task description. It supports custom automation scenarios across projects by providing configurable worktree handling and interactive prompts for task specification.

What's the best way to keep automated repository experiments isolated from the main codebase?

The best way to isolate automated experiments is spawning a worker agent into a dedicated git worktree. This approach deploys the agent flexibly across an isolated worktree rather than the main repository, keeping your primary codebase clean.

Does spawning a Claude agent require special role assignment for repository automation?

Spawning a Claude agent for repository automation includes role assignment as part of its configurable handling. You specify the task through interactive prompts, and the agent is assigned a role to execute research, code review, or custom tasks.

Can I use a spawned agent for both research investigations and PR commentary?

Yes, a spawned Claude agent is applicable for both research investigations and PR commentary. It handles multiple repository automation tasks including code review, PR comments, and custom tasks within a configured worktree environment.