vibeeval

vibeeval / vibecosystem

Public

119 agents, 202 skills, 48 hooks — AI software team built on Claude Code. No custom model. No custom API. Just good engineering.

ai-agents ai-software-team automation claude-code claude-skills
58
1
100% credibility
Found Mar 20, 2026 at 59 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

vibecosystem transforms a single AI coding assistant into a self-managing team of 119 specialized agents that handle planning, development, review, testing, and learning from errors with zero setup required.

How It Works

1
🔍 Discover vibecosystem

You hear about a helpful tool that turns your single AI coding helper into a full team of smart assistants.

2
📥 Get it ready

Download it from the sharing site and run a quick setup script that adds everything to your helper.

3
💻 Open your project

Go to your coding folder and chat with your AI helper just like you always do.

4
✨ Team comes alive

Say 'add a new feature' and suddenly a group of specialized helpers starts planning, building, checking, and improving your code all by themselves.

5
🔧 Watch it work

Your code gets written, tested for mistakes, reviewed for quality, and even learns from any slip-ups to do better next time.

🎉 Coding feels effortless

Now every project gets handled by your tireless AI team, saving you time and delivering polished results you can trust.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 59 to 58 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 vibecosystem?

vibecosystem transforms Claude Code into a self-organizing AI software team with 119 agents, 202 skills, and 48 hooks, all built in TypeScript without custom models or APIs—just good engineering. It solves the single-assistant bottleneck by coordinating specialized roles for planning, coding, reviewing, testing, and error learning across phases. Install via a one-line script into ~/.claude/, then prompt normally to activate the swarm.

Why is it gaining traction?

No config needed post-install, yet it delivers agent swarms, dev-QA loops, and cross-project learning that evolve rules from mistakes—features rivals like Cursor or aider lack. On GitHub (vibeeval/vibecosystem), the 119 agents hook devs tired of manual coordination, with self-improving pipelines sharing patterns globally. Termux users praise the lightweight, local setup.

Who should use this?

Claude Code power users building APIs, web apps, or prototypes solo—think backend devs handling full-stack tasks or indie hackers scaling MVPs. Skip if you're not deep into Anthropic tools; perfect for those debugging complex refactors or TDD workflows without a human team.

Verdict

Grab it if Claude Code is your daily driver—the 119 agents deliver real team vibes fast. But 58 stars and 1.0% credibility signal early-stage risks; test on non-critical projects first. Strong docs and MIT license make it a low-risk experiment.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.