The curated shortlist of repos that power AI agents, MCP servers, and Web3 tools
Turn what you learned into a concrete stack decision.
01AI Agent ToolboxBuild autonomous AI agents that can read, write, and actWant the shortlist in your inbox?
Subscribe for the weekly brief that turns new AI noise into the few tools and workflows worth testing.
Curated bundles that help you move from this guide into a working stack.
The essential toolkit for building AI agents with the Model Context Protocol. This pack includes MCP servers, agent frameworks, and workflow orchestration tools that let AI agents interact with real systems from GitHub to databases to Slack.
Open pack →A curated collection of essential tools for DeFi development. From smart contract frameworks to Web3 libraries, this pack covers the full stack for building decentralized finance applications on Ethereum and EVM chains.
Open pack →A complete starter kit for Solana development. This pack combines Web3 libraries, AI agent SDKs for Solana, trading bots, and learning resources to get you from zero to mainnet deployment fast.
Guide
The AI Builder's App Stack
Every app and tool you need to go from idea to deployed project
Guide
OpenClaw + Claude Code: Commands Cheat Sheet
Every command, shortcut, and workflow tip for both tools in one place
Guide
How to Build an AI Agent That Makes Money
From toy project to revenue — practical monetization patterns for autonomous AI agents
There are thousands of AI repos on GitHub. Most are demos, abandoned experiments, or forks of forks. This list is the 20 we actually use and recommend — repos that are actively maintained, well-documented, and solve real problems.
Organized by what you're building.
Model Context Protocol (MCP) servers let AI tools interact with external systems. These are the ones worth installing.
The official collection of MCP servers from Anthropic. Includes servers for GitHub, filesystem, PostgreSQL, Slack, Google Drive, and more. Start here — it's the reference implementation.
Community-curated list of every MCP server worth knowing about. Organized by category (databases, APIs, DevOps, AI). Use this to find servers for specific integrations.
Browser automation as an MCP server. Let your AI agent navigate websites, fill forms, take screenshots, and extract data. Essential for scraping and testing workflows.
Web search capability for AI agents via Brave Search API. Free tier (2,000 searches/month) is enough for personal agent use. Better privacy than Google alternatives.
Library documentation lookup as an MCP tool. Instead of your AI hallucinating API details, Context7 fetches real, up-to-date documentation. Dramatically reduces coding errors.
Tools for building multi-step, autonomous AI agents.
Build stateful, multi-step AI agents as graphs. Each node is an LLM call or action, edges define the flow. The most popular framework for complex agent workflows. Python-first.
Multi-agent orchestration where agents have roles, goals, and backstories. Great for simulating teams — a "researcher" agent feeds a "writer" agent feeds a "reviewer" agent. Python.
Microsoft's multi-agent conversation framework. Agents talk to each other to solve problems. Strong at complex reasoning tasks where one model's output feeds another's input.
Autonomous research agent that plans searches, gathers sources, evaluates credibility, and writes comprehensive reports with citations. Impressive out-of-the-box for deep research tasks.
The best TypeScript SDK for building AI-powered apps. Unified interface for Claude, GPT, Gemini, and open-source models. Streaming, tool use, structured output — all handled cleanly.
Repos designed to teach, not just to use.
Microsoft's step-by-step guide to the Model Context Protocol. 25 lessons covering everything from "what is MCP?" to building production servers. The best free MCP course available.
Learn Solidity (Ethereum smart contract language) through 50+ practical lessons. Each lesson is a self-contained concept with code examples. The most accessible Solidity tutorial.
smartcontractkit/full-blockchain-solidity-course-js
Patrick Collins' legendary free course covering Solidity from zero to production. 32+ hours of content with hands-on projects. The gold standard for blockchain development education.
codecrafters-io/build-your-own-x
Not AI-specific, but essential. Tutorials for building databases, compilers, Docker, Git, and more from scratch. Understanding how tools work under the hood makes you a better AI-assisted coder.
Repos that improve your day-to-day development workflow.
Community-maintained collection of .cursorrules files — instructions that configure how Cursor AI behaves for specific frameworks (Next.js, Python, Rust, etc.). Copy the right rules for your stack and Cursor becomes dramatically more useful.
Official Claude Code repository with documentation, examples, and community contributions. Bookmark this for the latest features, MCP integrations, and best practices.
Not an AI tool, but the most AI-friendly UI component library. Copy-paste components (not a dependency). Claude Code and Cursor generate excellent shadcn/ui code because it's in their training data.
Repos at the intersection of AI agents and blockchain.
AI agent toolkit for Solana — wallet management, token swaps, NFT operations, and DeFi interactions. The reference implementation for building AI agents that interact with Solana.
Security-focused repo that reproduces real DeFi exploits. Not just for security researchers — understanding how hacks work makes you build more secure smart contracts. Essential before deploying anything with real money.
Multi-agent simulation framework with built-in support for crypto trading, social media, and autonomous decision-making. Popular for building AI agents that operate in the crypto ecosystem.
Don't try to learn all 20 at once. Pick based on what you're building:
Building an AI coding tool? → Start with #1, #5, #10, #11 Building AI agents? → Start with #6, #7, #9, #11 Building in Web3/DeFi? → Start with #12, #13, #18, #19 Just getting started? → Star them all, then start with #11, #14, #15
Pro tip: Use GitHub's "Star" feature as a bookmark. Star repos now, explore them when you need them. Stars also help the maintainers — it signals that their work matters.
Want a curated toolkit instead of individual repos? Check out our Starter Packs — pre-assembled collections of tools for specific use cases.