jnMetaCode

jnMetaCode / agency-orchestrator

Public

YAML-first multi-agent orchestrator | 不用写代码的 AI 团队 — 186 个中文角色，零代码编排，支持 DeepSeek/Claude/OpenAI/Ollama

ai-agents chinese claude crewai-alternative deepseek
19
2
100% credibility
Found Mar 23, 2026 at 19 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Agency Orchestrator is a workflow tool that coordinates teams of predefined AI agents to collaboratively complete tasks such as product reviews, content creation, or story generation using simple descriptive files.

How It Works

1
📰 Discover the AI Team Builder

You hear about Agency Orchestrator, a simple way to make teams of expert AI helpers collaborate on real projects like writing stories or reviewing ideas.

2
📦 Download Expert Personalities

Grab the ready pack of 186 specialized AI roles, from product managers to storytellers and designers, all set up for you.

3
💻 Open Your AI Chat Helper

Launch your friendly AI assistant in a coding playground like Claude or Cursor where it can run team projects.

4
🔮 Add Team Superpower

Give your AI a special skill to understand and run collaborative workflows, no setup hassle.

5
🎯 Pick a Ready Project

Choose an example like 'create a short story' or 'review a product idea' and share your starting idea.

6
🤝 Watch the Team Collaborate

Your AI roles jump in one by one or together, passing ideas, researching, and building the result step by step—it feels magical.

🎉 Enjoy Perfect Results

Get a folder of polished outputs like a full story or detailed review, ready to use or share.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 19 to 19 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is agency-orchestrator?

Agency-orchestrator is a TypeScript-based, YAML-first multi-agent orchestrator that assembles AI teams from 186 pre-built Chinese roles like product managers, engineers, and marketers. You define workflows in simple YAML files specifying steps, tasks with variable passing, and dependencies; it auto-builds a DAG for sequential or parallel execution across DeepSeek/Claude/OpenAI/Ollama. Developers get zero-code pipelines for tasks like product reviews or story creation, runnable via CLI commands like `ao run workflow.yaml` or integrated into tools like Claude Code without API keys.

Why is it gaining traction?

It skips Python boilerplate from CrewAI or LangGraph, delivering 186 ready roles and native LLM connectors in a lightweight npm package with just two dependencies. CLI tools like `ao plan` visualize DAGs, `ao validate` checks syntax, and outputs track tokens/duration—users notice instant parallelism and retries without setup hassle. The hook: download roles with `ao init`, tweak YAML, and ship collaborative AI output.

Who should use this?

Product leads automating PRD reviews with parallel tech/UX agent feedback. Content creators chaining research-to-draft pipelines for platforms like WeChat or Xiaohongshu. Devs prototyping agency setups in Cursor or CLI for batch tasks, especially needing Chinese-tuned prompts.

Verdict

Grab it for fast multi-agent experiments—solid docs and templates make it approachable despite 19 stars and 1.0% credibility score signaling early maturity. Test thoroughly before production; roadmap promises UI and loops soon.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.