kobe

Manage parallel AI coding agent sessions using git worktrees and PTY processes.

94|7|Updated May 9, 2026
One-click install
npx skills add https://github.com/Sma1lboy/kobe --skill kobe-sma1lboy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kobe
Source: https://github.com/Sma1lboy/kobe/tree/main/.agents/skills/kobe
Command: npx skills add https://github.com/Sma1lboy/kobe --skill kobe-sma1lboy

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill solves the challenge of managing multiple, isolated AI coding agent sessions, preventing task interference and ensuring work persists across disconnections.

Core Features & Use Cases

  • Parallel Task Management: Run multiple coding agents simultaneously, each in its own git worktree and branch.
  • Session Persistence: Maintain active agent sessions that survive terminal disconnections or SSH drops.
  • Workflow Automation: Fan out complex prompts across different agents, supervise progress, and merge the best results.

Quick Start

Use the kobe skill to create a new task with a specific prompt and vendor engine for your current repository.

Frequently Asked Questions about kobe

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

FAQPage Schema
How do I run parallel AI coding agents without task interference?

To run parallel AI coding agents without task interference, you can orchestrate isolated sessions using multiplexed git worktrees. This ensures each agent operates on its own branch while maintaining separate PTY-based engine processes and session state.

How do I keep AI agent sessions active after an SSH disconnect?

To keep AI agent sessions active after an SSH disconnect, you need a daemon-backed environment that maintains session state. This approach handles inter-process communication and ensures your coding agent workflows persist across terminal drops.

Can I fan out complex prompts across different AI agent vendors?

Yes, you can fan out complex prompts across different AI agent vendors. The orchestration task supports multiple agent vendors, allowing you to supervise progress, compare automated reporting, and merge the best results from parallel sessions into a single branch.

What is the best way to automate branch merging for multiple coding agents?

The best way to automate branch merging for multiple coding agents is to use a session multiplexer that supervises parallel task progress. It orchestrates the lifecycle of agent sessions and merges the best results from isolated git worktrees automatically.

Do I need a daemon to manage parallel coding agent workflows?

Yes, you need a daemon-backed environment to manage parallel coding agent workflows. The daemon is required to maintain session state and handle the inter-process communication necessary for task orchestration and PTY-based engine processes.

Why use git worktrees for isolated AI coding agent sessions?

Git worktrees are used for isolated AI coding agent sessions to prevent task interference by providing separate working directories and branches. This multiplexing approach allows multiple agents to operate simultaneously within the same repository.