explaingit

kky42/agent_hub

Analysis updated 2026-05-18

★ 1PythonAudience · developerComplexity · 2/5Setup · easy

TLDR

A personal repository template for managing and syncing installable skills across multiple AI coding agents and computers.

Mindmap

mindmap
  root((agent_hub))
    What it does
      Manages agent skills
      Syncs across machines
      Symlinks personal skills
    Tech stack
      Bash
      YAML
    Use cases
      Single source of truth
      Install third party skills
      Sync new machine
    Audience
      Developers
      Multi agent users

Code map

Detail Auto

An interactive map of this repo's files and how they connect — its source is parsed live in your browser. Click Visualize to build it.

filefunction / class

What do people build with it?

USE CASE 1

Keep a single source of truth for personal AI agent skills across machines.

USE CASE 2

Install third-party skills from a YAML manifest into multiple agents at once.

USE CASE 3

Sync skills to a new computer by pulling the repo and running one script.

What is it built with?

BashYAML

How does it compare?

kky42/agent_huba-bissell/unleash-liteabhiinnovates/whatsapp-hr-assistant
Stars111
LanguagePythonPythonPython
Setup difficultyeasyhardhard
Complexity2/54/53/5
Audiencedeveloperresearcherdeveloper

Figures from each repo's GitHub metadata at analysis time.

How do you get it running?

Difficulty · easy Time to first run · 5min

In plain English

agent_hub is a personal repository template for managing "skills", installable capability packages, across multiple AI coding agents such as Codex, Pi, and Claude Code, and keeping those skills synchronized across multiple computers. The problem it addresses is that when you use several AI coding agents on different machines, managing which skills are installed on each becomes tedious. agent_hub provides a single source of truth: a skills/ directory for personal skills you write yourself, and a thirdparty-skills.yml file that lists skills from external sources you want installed. Running ./scripts/skill-sync is the main command. It reads both the personal skills folder and the YAML manifest, then creates symbolic links from your personal skill folders into each agent's runtime directory (for example, ~/.agents/skills/ and ~/.claude/skills/), and calls npx skills to install third-party entries. The command is safe to run repeatedly. Because personal skills are linked rather than copied, edits to a skill are immediately visible to the agent without re-running sync. Third-party skills are declared in a YAML file specifying the skill name, source repository, and the agents to install it for. Removing a skill means deleting its YAML entry and re-running skill-sync. If an upstream source disappears, the sync warns and continues with the rest. To keep skills consistent across machines, commit and push changes to the repository, then on another computer run git pull followed by skill-sync. The target directories for personal skills can be customised using the AGENT_HUB_SKILL_TARGETS environment variable.

Copy-paste prompts

Prompt 1
Run skill-sync in agent_hub and explain what it linked into my agent runtime directories.
Prompt 2
Add a new third-party skill entry to thirdparty-skills.yml and sync it.
Prompt 3
Set the AGENT_HUB_SKILL_TARGETS environment variable to change where personal skills are linked.

Frequently asked questions

What is agent_hub?

A personal repository template for managing and syncing installable skills across multiple AI coding agents and computers.

What language is agent_hub written in?

Mainly Python. The stack also includes Bash, YAML.

How hard is agent_hub to set up?

Setup difficulty is rated easy, with roughly 5min to a first successful run.

Who is agent_hub for?

Mainly developer.

Open on GitHub → Explain another repo

This repo across BitVibe Labs

Verify against the repo before relying on details.

Alternatives & similar repos