Can I use different AI models for the spawned agents?

Yes, the skill supports various agent types including Claude, Aider, Codex, and Gemini through the --with-agent flag.

How do I see what the spawned agents have accomplished?

You can use 'agent-cli dev status' to list all active worktrees and 'agent-cli dev run cat .claude/REPORT.md' to review the conclusion report written by the agent.

Why is --prompt-file recommended over --prompt?

The --prompt-file option avoids shell quoting issues with special characters like exclamation marks or backticks, making it much more reliable for complex, multi-line instructions.

How does this skill prevent Git conflicts?

It utilizes Git worktrees to create entirely separate directories for each branch, ensuring that agents work in isolation without overlapping local file changes until you are ready to merge.

Does it work with existing branches?

Yes, you can use the --from HEAD flag to spawn an agent to fix, test, or review code based on your current working branch.

Parallel AI Agent Orchestrator

Name: Parallel AI Agent Orchestrator
Author: basnijholt

bybasnijholt

•

103

•

Productivity & Workflow

Spawns and manages multiple AI coding agents in isolated Git worktrees for parallel development and task delegation.

The Parallel AI Agent Orchestrator skill empowers Claude to scale development by delegating tasks to independent AI agents operating in isolated Git worktrees. By leveraging the agent-cli tool, it automates the creation of new branches, handles project setup, and initializes specialized agents for specific features or refactors. This is particularly useful for complex projects where multiple independent components can be developed simultaneously without the risk of merge conflicts or environment contamination, allowing for massive productivity gains through multi-agent orchestration.

Key Features

01Isolated Git worktree creation for conflict-free parallel development

02103 GitHub stars

03Automated reporting and summary generation for completed tasks

04Automatic project setup and dependency installation for new agents

05Support for multi-line complex instructions via prompt files

06Centralized status monitoring for all active AI worktrees

Use Cases

01Performing large-scale refactors by splitting work across modules

02Executing Test-Driven Development by assigning tests and implementation to separate agents

03Developing independent features like Authentication and API integrations in parallel

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add basnijholt/agent-cli agent-cli-dev

For use in Claude.ai and ChatGPT

Download Skill