Top AI Agent, MCP & Coding Skills GitHub Repositories (2026)

Autonomous AI agents have moved from experimental prompt loops to the fastest-growing category across all of GitHub. While our Top 100 Most Starred GitHub Repositories leaderboard tracks the broader open-source ecosystem, engineering leaders increasingly need a focused index of the specific building blocks that power modern agentic systems.
This companion index isolates 50+ high-signal AI agent repositories across five architectural tracks: Coding Agents & Harnesses, Agent Skills & Prompts, Workflows, MCP & Orchestration, Memory & Knowledge Graphs, and Security & Scientific Agents.
Quick Snapshot: Open-Source Agent Stack in 2026
- Coding Harnesses & Terminal Agents: rea, ECC, deepseek-harness, opencode, claude-code, openai/codex, and orca lead developer adoption for autonomous code generation, review, and multi-agent execution.
- Composable Skills (
SKILL.md): skills (mattpocock/skillsandanthropics/skills), superpowers, ponytail, caveman, and agent-skills standardise reusable engineering playbooks and token compression. - Orchestration & MCP Ecosystem: openclaw, n8n, dify, langflow, langchain, browser-use, and modelcontextprotocol/servers connect reasoning models to enterprise APIs, browsers, and databases.
- Persistent Memory & Codebase Graphs: graphify, claude-mem, Understand-Anything, hindsight, and codebase-memory-mcp give agents deterministic AST structural awareness and cross-session memory.
Interactive Index: Top AI Agent, MCP & Skills Repositories
Filter the index below by architectural track or search by language, protocol, or capability (MCP, memory, security, TypeScript, Python, Rust).
| # | Repository & Architecture Role | Language | Stars | Forks |
|---|---|---|---|---|
| #1 | Reverse engineer anything with agents, from app behavior down to native binaries. | TypeScript | 451,346 | 111.1k |
| #2 | The AI that really does things. Any OS. Any Platform. The lobster way. 🦞 | TypeScript | 394,324 | 82.9k |
| #3 | Skills for Real Engineers. Straight from my .agents directory. | Shell | 321,460 | 26.9k |
| #4 | An agentic skills framework & software development methodology that works. | Shell | 315,996 | 28.1k |
| #5 | The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond. | JavaScript | 301,019 | 44.4k |
| #6 | DeepSeek Harness: Everything is a Plugin. | TypeScript | 278,646 | 34.3k |
| #7 | The agent that grows with you | Python | 262,597 | 58.8k |
| #8 | A single CLAUDE.md file to improve Claude Code behavior, derived from Andrej Karpathy's observations on LLM coding pitfalls. | Universal | 232,049 | 23.1k |
| #9 | The open source coding agent. | TypeScript | 226,025 | 30.5k |
| #10 | Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations. | TypeScript | 210,133 | 61.5k |
| #11 | Supercharge your AI agents with data from the web and beyond. Building the library for superintelligence. 🔥 | TypeScript | 200,719 | 10.6k |
| #12 | An agent-managed museum exhibit, built in Rust with Gajae-Code / LazyCodex — developed and maintained with no human intervention. | Rust | 194,796 | 107.4k |
| #13 | Public repository for Agent Skills | Python | 192,785 | 22.7k |
| #14 | Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote. | JavaScript | 188,726 | 10.5k |
| #15 | AutoGPT is the vision of accessible AI for everyone, to use and to build on. Our mission is to provide the tools, so that you can focus on what matters. | Python | 187,991 | 45.9k |
| #16 | f.k.a. Awesome ChatGPT Prompts. Share, discover, and collect prompts from the community. Free and open source — self-host for your organization with complete privacy. | MDX | 176,887 | 22.6k |
| #17 | A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables. | Shell | 167,721 | 27.2k |
| #18 | Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype to production without rebuilding the stack. | Python | 160,962 | 25.5k |
| #19 | Langflow is a powerful tool for building and deploying AI-powered agents and workflows. | Python | 156,523 | 10.4k |
| #20 | Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands. | TypeScript | 155,253 | 24.9k |
| #21 | A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode,DeepSeek Harness and more. Only official website: ccswitch.io | Rust | 153,560 | 10.1k |
| #22 | The agent engineering platform. | Python | 149,345 | 25.1k |
| #23 | 💫 Toolkit to help you get started with SDD or any other process! | Python | 149,284 | 13.4k |
| #24 | An AI skill that provides design intelligence for building professional UI/UX across multiple platforms. | Python | 148,013 | 15.7k |
| #25 | Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store. | Python | 145,886 | 14.0k |
| #26 | FULL Augment Code, Claude Code, Cluely, CodeBuddy, Comet, Cursor, Devin AI, Junie, Kiro, Leap.new, Lovable, Manus, NotionAI, Orchids.app, Perplexity, Poke, Qoder, Replit, Same.dev, Trae, Traycer AI, VSCode Agent, Warp.dev, Windsurf, Xcode, Z.ai Code, Dia & v0. (And other Open Sourced) System Prompts, Internal Tools & AI Models | Universal | 144,610 | 34.7k |
| #27 | Use Garry Tan's exact Claude Code setup: 23 opinionated tools that serve as CEO, Designer, Eng Manager, Release Manager, Doc Engineer, and QA | TypeScript | 141,170 | 20.8k |
| #28 | Lightweight coding agent that runs in your terminal | Rust | 134,855 | 21.4k |
| #29 | A collection of DESIGN.md files analysis by popular brand design systems. Drop one into your project and let coding agents generate a matching UI. | Universal | 126,690 | 14.0k |
| #30 | TradingAgents: Multi-Agents LLM Financial Trading Framework | Python | 121,963 | 23.3k |
| #31 | Orca lets you run a fleet of coding agents (e.g. Claude Code, Codex) with your own subscription. Available on desktop, mobile and remote runtime. | TypeScript | 121,915 | 7.7k |
| #32 | AI agent toolkit: unified LLM API, agent loop, TUI, coding agent CLI | TypeScript | 121,380 | 15.6k |
| #33 | Open-source Python library for AI browser agents, plus Browser Use Cloud: stealth browsers and a hosted agent API. | Python | 120,671 | 13.5k |
| #34 | 🪨 why use many token when few token do trick. Viral skill + proxy for coding agents that cuts 65% of tokens by talking like a caveman. | Go | 119,002 | 6.9k |
| #35 | Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + More | TypeScript | 116,027 | 10.1k |
| #36 | Production-grade engineering skills for AI coding agents. | JavaScript | 115,528 | 12.1k |
| #37 | The open-source app everyone uses to manage agents at work | TypeScript | 113,055 | 18.8k |
| #38 | Graphs that teach > graphs that impress. Turn any code into an interactive knowledge graph you can explore, search, and ask questions about. Works with Claude Code, Codex, Cursor, Copilot, Gemini CLI, and more. | TypeScript | 91,985 | 7.8k |
| #39 | Model Context Protocol Reference Servers | TypeScript | 91,921 | 11.9k |
| #40 | A Simple and Universal Swarm Intelligence Engine, Predicting Anything. 简洁通用的群体智能引擎,预测万物 | Python | 85,258 | 12.9k |
| #41 | World's first open-source, agentic video production system. 12 production pipelines, 100+ tools, 700+ agent skill and production-knowledge files. Turn your AI coding assistant into a full video production studio. | Python | 79,558 | 10.6k |
| #42 | A coding-agent skill for multi-phase security audits with independently verified, machine-readable findings | JavaScript | 69,813 | 4.6k |
| #43 | Editorial diagram design for Claude Code, Codex, GitHub Copilot, Cursor, Factory Droid, Qoder, and Pi. 45 diagram types. Self-contained HTML + SVG. No shadows. No Mermaid slop. | HTML | 66,433 | 4.4k |
| #44 | Hindsight: Agent Memory That Learns | Python | 60,412 | 7.4k |
| #45 | Turn any AI agent into an AI Scientist. The #1 Agent Skills library for science, used by 250,000+ scientists worldwide. 177 ready-to-use validated skills plus 100+ scientific databases covering biology, chemistry, medicine, and drug discovery. Compatible with Cursor, Claude Code, Codex, Pi, Antigravity, and the open Agent Skills standard. | Python | 56,802 | 5.0k |
| #46 | Anti-AI-slop design skill for Claude Code, Cursor, and Codex. | CSS | 52,998 | 2.8k |
| #47 | Local code intelligence for developers and AI agents. Connect code, docs and infrastructure in a persistent graph, with CBM Lens, MCP tools, agent integrations and change-aware test selection. | C | 48,480 | 4.0k |
| #48 | GitNexus: The Zero-Server Code Intelligence Engine | TypeScript | 48,422 | 5.2k |
| #49 | The fastest, cheapest, most efficient open-source agent harness. Run more than 500 agents on a $10 VPS. | Rust | 42,647 | 4.2k |
| #50 | 📨 The ultimate agentic social media scheduling tool 🤖 | TypeScript | 41,337 | 7.9k |
| #51 | One shared memory for every coding agent on your team. Claude Code, Cursor and Codex get a map of your code and your team's past decisions, so they pick up where you left off. | TypeScript | 11,199 | 999 |
| #52 | claude-red is a curated library of offensive security skills designed for the Claude skills system. Each skill is a structured SKILL.md file that primes Claude with expert-level methodology for a specific attack surface — from SQLi to shellcode, EDR evasion to exploit development. | Python | 10,588 | 1.4k |
| #53 | Sellbird (formerly Harvey): the open-source outreach platform that runs on your Claude or ChatGPT plan. Finds companies and decision makers, verifies emails, writes personal cold email, sends it and handles the replies. | Python | 5,113 | 1.1k |
| #54 | This Repository is a collection of different ethical hacking tools and malware's for penetration testing and research purpose written in python, ruby, rust, c++, go and c. | C | 1,100 | 282 |
| #55 | Free Active Directory pentesting tool for Linux, macOS and Windows. Automates AD and LDAP enumeration, Kerberoasting, ASREPRoast, password spraying, ADCS/ESC1, DCSync, RBCD, gMSA, shadow credentials, NTLM relay and credential dumping across 104 techniques, mapping BloodHound-compatible attack paths to Domain Admin. NIS2 / ISO 27001 reports. | Python | 787 | 107 |
Choosing the Right Layer for Enterprise Agent Workloads
Open-source star counts measure developer enthusiasm, not enterprise readiness. When transitioning from local CLI coding assistants to customer-facing or internal business automation, separate your architecture into four distinct decisions:
1. Developer Harnesses vs Production Workflow Engines
Terminal agents and skill packs (claude-code, opencode, codex, superpowers, ECC) excel on developer workstations where an engineer reviews diffs before merge. For unattended backend workflows—such as invoice triage, service-desk routing, or document extraction—pair structured orchestrators (dify, langflow, n8n, or Azure Foundry Agent Service) with explicit retry budgets, dead-letter queues, and deterministic schema validation. See our blueprint on Agentic AI on Azure for the 5-layer production reference architecture.
2. Model Context Protocol (MCP) & Scoped Tool Boundaries
Standardising tool connectors around modelcontextprotocol/servers prevents vendor lock-in when switching frontier models. However, community MCP servers frequently request broad API tokens or filesystem permissions. Always wrap third-party MCP servers behind least-privilege service principals, network egress controls, and read/write separation.
3. Deterministic AST Graphs Over Pure Vector RAG
A standout 2026 shift is the rise of graphify, Understand-Anything, GitNexus, and codebase-memory-mcp. Pure vector embeddings struggle with call-graph dependencies, SQL foreign keys, and cross-service contracts; deterministic AST knowledge graphs give agents exact dependency paths at a fraction of the token cost.
4. Supply-Chain & Vulnerability Gateways
Fast-moving AI repositories often pull dozens of transitive Python and npm packages. Before deploying any open-source agent framework, MCP server, or containerised worker into your cloud environment:
- Scan Lockfiles & Container Images: Run your
Dockerfile,requirements.txt, orpackage-lock.jsonthrough our free Trivy Vulnerability & Dockerfile Security Scanner to catch critical CVEs and exposed secrets prior to CI/CD build. - Assess Baseline Cyber Posture: Benchmark identity, patching, and execution controls against Australian compliance baselines using our Cyber Risk Assessment Tool.
Designing a governed multi-agent workflow, custom MCP integration, or private enterprise AI platform in Australia? Explore our AI Agents & Workflow Automation engineering practice or book a technical architecture session with Gurinder Singh.
Frequently Asked Questions
Direct engineering answers to key technical and migration considerations.
Written by Gurinder Singh
AuthorPrincipal Cloud & Software Architect at AgenorIT. Specialising in Microsoft Azure Landing Zones, Microsoft Entra identity architectures, Microsoft Fabric lakehouses, and high-performance digital products for Australian organisations.
Related Architecture & Engineering Insights
Top 100 Most Starred GitHub Repositories (2026 Guide)
The definitive curated index of the top 100 most starred GitHub repositories in 2026, featuring star counts, forks, tech stacks, and direct source code links.
Agentic AI on Azure: Building Production-Ready AI Agents
Build, secure, and evaluate production AI agents on Microsoft Azure. Architectural patterns, Microsoft Foundry, Entra Agent ID, and Fabric integration.
Need Senior Architecture Guidance on Your Platform?
Speak directly with an experienced engineer about cloud infrastructure, data pipelines, or software development.