oribarilan

oribarilan / 97

Public

Agent skills distilled from the hard-won lessons of world-renowned programmers, in the spirit of "97 Things Every Programmer Should Know"

agent-skills ai-agents best-practices claude-code claude-code-plugin
15
0
100% credibility
Found May 08, 2026 at 15 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
JavaScript
AI Summary

97 is a collection of skills for AI coding assistants that automatically applies best practices from expert programming books during tasks like refactoring, testing, API design, and deployment.

How It Works

1
🔍 Find 97 Helper

You stumble upon 97, a smart add-on that makes your AI coding buddy follow expert programming wisdom.

2
Pick Your AI Buddy
🧑‍💻
Claude Code

Tell it to grab 97 from the shared collection.

💻
Copilot Chat

Add 97 through its collection of helpers.

⚙️
OpenCode

Note it in your setup notes and refresh.

3
✅ Set It Up Quick

Follow a simple step or two in your chosen tool to welcome 97's skills aboard.

4
✨ Dive into Coding

Fire up your project and chat with your AI about writing code, fixing issues, or adding tests – now supercharged with pro tips!

5
💡 Smart Guidance Kicks In

Your AI pauses at key moments to apply time-tested rules for clean names, safe handling, solid tests, and more.

🎉 Masterful Code Achieved

Step back and admire your polished, reliable software that feels professional and built to last.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 15 to 15 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 97?

97 is a JavaScript plugin pack that supercharges AI coding agents like Claude Code, GitHub Copilot CLI, and OpenCode with 12 trigger-based skills drawn from classics like "97 Things Every Programmer Should Know." It kicks in automatically during refactors, API design, testing, error handling, security checks, and commits, applying principles from Fowler, Liskov, and Ousterhout to enforce clean code, observability, and deploy hygiene. Developers get agent skills claude and agent github copilot that cite sources on-demand, turning vague AI outputs into principled code without manual prompts.

Why is it gaining traction?

Unlike process-focused plugins like superpowers or role-based ones like BMAD, 97 targets craft: what makes code good, from SRP in functions to parse-don't-validate in trust boundaries. The hook is its moment-of-decision triggers—before-you-refactor or pre-commit-self-review—that weave agent skills examples into daily workflows, with full attribution for learning. It complements agent github copilot vscode or cli setups, standing out for language-agnostic wisdom over generic boosts.

Who should use this?

Backend devs building APIs with GitHub Copilot CLI who hate flaky tests or security oversights; full-stack teams on Claude Code needing observability in production handlers; OpenCode users tackling domain modeling or deploys. Ideal for solo devs or pairs using agent skills anthropic tools daily, especially if you're evaluating agent skills github for consistent quality.

Verdict

Try it if you're deep in agent github code workflows—installs in one command per harness, solid docs, and zero runtime deps make experimentation low-risk. At 15 stars and 1.0% credibility, it's early beta with thin adoption, but the curated principles deliver immediate wins over raw AI.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.