skills.rest
hbvg234
hbvg234COMMUNITY
Explore all skills by this author →
GitHub

agenthub

Coordinate parallel AI agents in isolated git worktrees and merge the best result.

1|Updated Apr 9, 2026
Search Tags:#collaboration#agent#evaluation#multi-agent#orchestration#git-worktree
One-click install
npx skills add https://github.com/hbvg234/jnmt.vn --skill agenthub-hbvg234
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agenthub
Source: https://github.com/hbvg234/jnmt.vn/tree/main/.claude/skills/agenthub
Command: npx skills add https://github.com/hbvg234/jnmt.vn --skill agenthub-hbvg234

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill enables orchestration of multiple AI agents to work in parallel on the same task, using isolated git worktrees to prevent interference and ensure fair evaluation.

Core Features & Use Cases

  • Multi-agent collaboration with parallel worktrees to explore diverse approaches
  • Evaluation via metrics or LLM judges, with winner merging and loser's archiving
  • Predefined agent templates (optimizer, refactorer, test-writer, bug-fixer) for rapid experimentation
  • Git-backed session management, message boards, and lifecycle orchestration for traceable results
  • Seamless integration with a local git repo workflow to reproduce and scale experiments

Quick Start

Initialize a session with a task, then spawn N agents to compete in isolated worktrees and merge the winner.

Frequently Asked Questions about agenthub

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

FAQPage Schema
How do I run multiple AI agents in parallel on the same task without them interfering?▼

You can coordinate parallel AI agents using isolated git worktrees. This approach spawns N agents to explore diverse solutions simultaneously while keeping their changes separate until a winner is merged.

How does multi-agent orchestration evaluate competing code variations?▼

Multi-agent orchestration evaluates competing code variations via metrics or LLM judges. It assesses each isolated branch's outcome, merges the winning branch, and archives the losing branches.

Can I use predefined agent templates for code optimization and bug fixing?▼

Yes, predefined agent templates are available for code optimization, refactoring, test writing, and bug fixing. These templates enable rapid parallel experimentation across isolated git worktrees.

What is the best way to manage multi-agent collaboration results in a local git repository?▼

The best way to manage multi-agent collaboration results is through git-backed session management. It tracks lifecycle orchestration, message boards, and final merges with proper tagging and cleanup for traceable results.

Does multi-agent orchestration work for research exploration and content generation?▼

Yes, multi-agent orchestration applies to research exploration and content generation. It benefits any task requiring multiple approaches explored in parallel, evaluating outcomes to select the best result.

Related Skills

Explore More Software Engineering Skills→
Moonsong-LabsMoonsong-Labs
3

using-git-worktrees

Create and validate isolated Git worktrees for feature development.

Official
Advanced
Veloxia-agencyVeloxia-agency

merge

Merge the winning agent branch into the base branch with no-ff merge and archive tags.

Community
Intermediate
Patasse97Patasse97

eval

Rank AgentHub session agent results using metrics or an LLM judge.

Community
Advanced
woxiangyangzhimaowoxiangyangzhimao
3

batch

Execute sequential code changes with checkpoints and rollback.

Community
Advanced