pejmanjohn

pejmanjohn / slot-machine

Public

An open-source skill for running parallel implementations, reviewing them independently, and selecting or synthesizing the best result.

ai-agents automation claude-code code-review codex
10
2
100% credibility
Found Mar 31, 2026 at 10 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Shell
AI Summary

A skill for AI coding environments that runs multiple independent AI agents on the same task to implement, review, judge, and deliver or combine the superior outcome.

How It Works

1
👀 Discover Slot Machine

You hear about a clever way to get the absolute best results from AI helpers by having them compete on the same job.

2
📥 Add to your toolkit

You simply copy this handy skill into your AI coding helper's collection, and it's ready to use.

3
💭 Describe your project

In your chat with the AI, you explain the coding feature or piece of writing you want created.

4
🎰 Pull the lever

You launch Slot Machine, and it fires up several AI workers who each tackle your task in their own unique style.

5
🔍 Helpers get reviewed

Independent checkers examine each worker's output, hunting for flaws and highlighting strengths with clear notes.

6
⚖️ Judge makes the call

A wise overseer compares everything side-by-side and picks the top performer or blends the finest pieces together.

🏆 Get winning results

Your project now shines with the highest-quality work—bugs fixed, ideas perfected, and ready to use.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 10 to 10 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 slot-machine?

Slot-machine is a Shell-based open-source skill for Claude Code that runs parallel AI agent implementations of the same spec, reviews them independently for bugs and quality, then picks the winner or synthesizes the best parts into superior code. Among github open source tools and open source developer skills, it tackles AI's probabilistic output—where one shot often yields bugs or suboptimal designs—by competing N attempts (like /slot-machine with 3 slots) across models like Claude or Codex, using git worktrees for isolation. Users get higher-quality code or writing with evidence-based decisions, plus artifacts like reviewer scorecards saved to disk.

Why is it gaining traction?

It stands out from basic parallel agents by enforcing blind reviews that catch crashes self-review misses, structured judging with pick/synthesize/reject verdicts, and seamless integration with skills like test-driven-development or codebase patterns. Developers hook on cross-model runs (Claude vs Codex finding different bugs) and profiles for coding/writing, turning vague "best-of-N" into reliable pipelines that boost test coverage 2x+ without manual intervention. As an open source github copilot alternative and open source skill management tool, its diversity hints and autonomous loop support make complex tasks reliable.

Who should use this?

Backend engineers implementing features with tradeoffs like robustness vs simplicity, AI workflow builders in Claude Code automating overnight loops, or teams shipping production code where bug costs exceed token spend. Ideal for specs clear enough for independent impls, like API handlers or schedulers, but skip mechanical edits. Full-stack devs exploring open source skills ontology for agentic coding will find it composes well with TDD or CE skills.

Verdict

Try it if you're deep in Claude Code and want bug-resistant AI codegen—strong docs, tests, and MIT license make setup trivial despite 10 stars and 1.0% credibility score signaling early maturity. Not production-ready for all stacks yet, but a smart bet for agent-heavy projects chasing open source skill matrix gains.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.