Agent teams

268 Expert Roles in One Repo, Free

The free GitHub roster that gives your AI a job title before you type a word

2min read
8numbered steps
7sections

What this actually is

Most people open a chat and spend the first five minutes describing who they wish they were talking to. "You are a senior product designer with ten years of experience." That brief is work, you rewrite it every time, and it is the reason two people get wildly different answers out of the same model.

This repo replaces that. It is a folder of 268 pre written agent definitions, sorted into 20 divisions. Each one already knows its role, its scope, and what it is supposed to hand back. You name the job. The right specialists pick themselves up.

The repo

github.com/jnMetaCode/agency-agents-zh

  • 18,585 stars
  • 268 agents
  • 20 divisions
  • Works across 18 tools

One note on provenance, because it matters if you go looking. This project has an English language parent at github.com/msitarzewski/agency-agents with a much larger star count and a slightly smaller roster. The counts above belong to the fork listed here, which is the one those numbers come from. Its README is in Chinese and 52 of its agents are built for Chinese platforms. Both are free. Pick whichever suits the market you build for.

The 20 divisions

Engineering, Design, Marketing, Legal, Product, Growth, Finance, Support, Research, Security, Data, Ops, Sales, Content, QA, Brand, Partners, People, Strategy, Studio.

Roughly 13 agents sit inside each one. That is the whole point: you are not picking a model, you are picking a role.

Free, for your email

Get the full setup steps + the copy-paste setup prompt

The rest of this guide (4 more sections) opens right here. The setup prompt, one paste into Claude that sets it all up, comes by email only. Use your real address.

Unlocks: Install it in under a minute · Where it works · The three habits that make it pay off · The honest limitation

50,000+ follow where these guides come from. No spam, reply “unsubscribe” any time.

Read next