I scanned 1,000+ repos and tested 200+ skills so you don't have to.
Here are the 90 AI tools that actually matter right now — skills, MCP servers, and GitHub repos in one list. Zero fluff.
Part 1: Claude Skills
There are 60,000+ community skills out there. It's an app store for AI coding knowledge. These 22 are the ones worth installing.
Document & Office (Official Anthropic) 📰
01 — PDF Processing — Read, extract tables, fill forms, merge/split. Highest-utility skill for knowledge workers. https://github.com/anthropics/skills/tree/main/skills/pdf
02 — DOCX — Create & edit Word docs with tracked changes, comments, formatting. https://github.com/anthropics/skills/tree/main/skills/docx
03 — PPTX — Slide decks from natural language. Layouts, charts, speaker notes. https://github.com/anthropics/skills/tree/main/skills/pptx
04 — XLSX — Formulas, analysis, charts via plain English. https://github.com/anthropics/skills/tree/main/skills/xlsx
05 — Doc Co-Authoring — Real collaborative writing. Human writes, Claude responds, back and forth. https://github.com/anthropics/skills/tree/main/skills/doc-coauthoring
Design & Creative 🎨
06 — Frontend Design — Escape "AI slop" aesthetics. Real design systems, bold typography. 277k+ installs. https://github.com/anthropics/skills/tree/main/skills/frontend-design
07 — Canvas Design — Social graphics, posters, covers — text in, PNG/PDF out. https://github.com/anthropics/skills/tree/main/skills/canvas-design
08 — Algorithmic Art — Fractal patterns, geometric compositions via p5.js. https://github.com/anthropics/skills/tree/main/skills/algorithmic-art
09 — Theme Factory — Batch-generate color schemes from one prompt. https://github.com/anthropics/skills/tree/main/skills/theme-factory
10 — Web Artifacts Builder — Calculators, dashboards via natural language. No frontend skills needed. https://github.com/anthropics/skills/tree/main/skills/web-artifacts-builder
Dev & Engineering 👨💻
11 — Superpowers — 20+ battle-tested skills. TDD, debugging, plan-to-execute pipeline. 96k+ stars. https://github.com/obra/superpowers
12 — Systematic Debugging — Root cause analysis first, fix second. 4-phase methodology. S9.2 on SkillHub. https://github.com/obra/superpowers
13 — File Search — Ripgrep + ast-grep mastery. S9.0 rated. https://github.com/massgen/massgen
14 — Context Optimization — Reduce token costs, KV-cache tricks. 13.9k stars. https://github.com/muratcankoylan/agent-skills-for-context-engineering
15 — Skill Creator (Official) — Meta-skill: describe a workflow, get a SKILL.md in 5 min. https://github.com/anthropics/skills/tree/main/skills/skill-creator
16 — Remotion Best Practices — AI video generation. 117k weekly installs. https://github.com/remotion-dev/remotion
Marketing & SEO 🗞️
17 — Marketing Skills by Corey Haines — 20+ skills: CRO, copywriting, SEO, email sequences, growth. https://github.com/coreyhaines31/marketingskills
18 — Claude SEO — Full-site audits, schema validation. 12 sub-skills. https://github.com/AgriciDaniel/claude-seo
19 — Brand Guidelines — Encode your brand into a skill. Auto-applies everywhere. https://github.com/anthropics/skills/tree/main/skills/brand-guidelines
Knowledge & Learning 🧠
20 — NotebookLM Integration — Claude + NotebookLM bridge. Summaries, mind maps, flashcards. https://github.com/PleasePrompto/notebooklm-skill
21 — Obsidian Skills — By Obsidian's CEO. Auto-tagging, auto-linking, vault-native. https://github.com/kepano/obsidian-skills
22 — Excel MCP Server — Manipulate Excel without Microsoft Excel. https://github.com/haris-musa/excel-mcp-server
All official skills: github.com/anthropics/skills
Browse 80k+ community skills: skillsmp.com
My Must-Have MCP Servers
Beyond skills — I love using these. Skills teach Claude HOW to do things. MCP gives it ACCESS to the outside world. These three changed my workflow.
Tavily
Search engine built for AI agents. Not blue links — clean structured data. Four tools: search, extract, crawl, map. Connects as remote MCP in one minute.
https://github.com/tavily-ai/tavily-mcp
Context7
Injects up-to-date library docs into your LLM's context. No more hallucinated APIs. No more deprecated methods. Just add "use context7" to your prompt. Supports Next.js, React, Supabase, MongoDB, thousands more.
https://github.com/upstash/context7
Task Master AI
Your AI's project manager. Feed a PRD → structured tasks with dependencies → Claude executes one by one. Turns chaotic sessions into a proper pipeline. Works across Cursor, Claude Code, Windsurf.
https://github.com/eyaltoledano/claude-task-master
Part 2: GitHub Repos
25 open-source engines powering the AI revolution.
Let's start with most famous one and go into more lowkey ones.
Agent Frameworks
23 — OpenClaw — The viral AI agent. Persistent, multi-channel, writes its own skills. 210k+ stars. https://github.com/openclaw/openclaw
24 — AutoGPT — Full agent platform for long-running tasks. https://github.com/Significant-Gravitas/AutoGPT
25 — LangGraph — Agents as graphs. Multi-agent orchestration. 26.8k stars. https://github.com/langchain-ai/langgraph
26 — OWL — Multi-agent cooperation. Tops GAIA benchmark. https://github.com/camel-ai/owl
27 — Dify — Open-source LLM app builder. Workflows, RAG, agents all-in-one. https://github.com/langgenius/dify
28 — CrewAI — Multi-agent with roles, goals, backstories. https://github.com/crewAIInc/crewAI
29 — CopilotKit — Embed AI copilots into React apps. https://github.com/CopilotKit/CopilotKit
Local AI
30 — Ollama — Run LLMs locally with one command. https://github.com/ollama/ollama
31 — Open WebUI — Self-hosted ChatGPT-like interface. https://github.com/open-webui/open-webui
32 — LlamaFile — LLM as single executable. Zero dependencies. https://github.com/Mozilla-Ocho/llamafile
33 — Unsloth — Fine-tune 2x faster, 70% less memory. https://github.com/unslothai/unsloth
Workflow & Automation
34 — n8n — Open-source automation, 400+ integrations + AI nodes. https://github.com/n8n-io/n8n
35 — Langflow — Visual drag-and-drop for agent pipelines. 140k stars. https://github.com/langflow-ai/langflow
36 — Huginn — Self-hosted web agents. Monitoring, alerts. Privacy-first. https://github.com/huginn/huginn
Search & Data
37 — GPT Researcher — Autonomous research → compiled reports. https://github.com/assafelovic/gpt-researcher
38 — Firecrawl — Any website → LLM-ready data. https://github.com/mendableai/firecrawl
39 — Vanna AI — Natural language → SQL. https://github.com/vanna-ai/vanna
Dev Tools
40 — Codebase Memory MCP — Codebase → persistent knowledge graph. https://github.com/DeusData/codebase-memory-mcp
41 — DSPy — Program (not prompt) foundation models. https://github.com/stanfordnlp/dspy
42 — Spec Kit (GitHub) — Spec-driven dev. Write specs, AI generates code. 50k+ stars. https://github.com/github/spec-kit
43 — NVIDIA NemoClaw — Secure sandbox for autonomous agents. https://github.com/NVIDIA/NemoClaw
Curated Collections
44 — Awesome Claude Skills — Best curated skill list. 22k+ stars. https://github.com/travisvn/awesome-claude-skills
45 — Anthropic Skills Repo — Official reference implementations. https://github.com/anthropics/skills
46 — Awesome Agents — 100+ open-source agent tools. https://github.com/kyrolabs/awesome-agents
47 — MAGI//ARCHIVE — Daily feed of fresh AI repos. https://tom-doerr.github.io/repo_posts/
40 fresh Github Repos Worth Watching
I scanned 1,000+ repos from Github forums and picked the most interesting AI projects. Don't forget to do secuirity check yourself.
Agent Orchestration & Multi-Agent 🤖
gstack — Claude Code as virtual engineering team https://github.com/garrytan/gstack
cmux — Multiple Claude agents in parallel https://github.com/craigsc/cmux
figaro — Orchestrate Claude agent fleets on desktop https://github.com/byt3bl33d3r/figaro
claude-squad — Terminal agents in parallel sessions https://github.com/smtg-ai/claude-squad
deer-flow (ByteDance) — Sub-agents and sandboxes through skills https://github.com/bytedance/deer-flow
SWE-AF — One API call spins up engineering team https://github.com/Agent-Field/SWE-AF
AIlice — Complex tasks → dynamic agents https://github.com/myshell-ai/AIlice
Agent Alchemy — Claude Code + plugins + task manager https://github.com/sequenzia/agent-alchemy
Agent Infrastructure & Security🔐
Ghost OS — AI agents operate every Mac app https://github.com/ghostwright/ghost-os
e2b/desktop — Isolated virtual desktops for agents https://github.com/e2b-dev/desktop
container-use (Dagger) — Containerized environments for coding agents https://github.com/dagger/container-use
Canopy — Encrypted P2P mesh for AI agents https://github.com/kwalus/Canopy
agent-governance-toolkit (Microsoft) — Security middleware for agents https://github.com/microsoft/agent-governance-toolkit
claude-code-security-review (Anthropic) — PRs analyzed for security https://github.com/anthropics/claude-code-security-review
promptfoo — Automated security testing for AI models https://github.com/promptfoo/promptfoo
Memory & Context📝
Mem9 — Memory system for AI agents https://github.com/mem9-ai/mem9
Codefire — Persistent memory for coding agents https://github.com/websitebutlers/codefire-app
Memobase — User profile memory for LLMs https://github.com/memodb-io/memobase
Coding Agents & Dev Tools💻
Qwen Code — Terminal AI coding agent by QwenLM https://github.com/QwenLM/qwen-code
gptme — Personal AI agent in terminal https://github.com/gptme/gptme
Claude Inspector — See hidden Claude Code prompt mechanics https://github.com/kangraemin/claude-inspector
TDD Guard — Enforces test-first for AI agents https://github.com/nizos/tdd-guard
rendergit (Karpathy) — Git repo → single file for humans and LLMs https://github.com/karpathy/rendergit
autoresearch (Karpathy) — Autonomous LLM training system https://github.com/karpathy/autoresearch
pydantic-ai — Type-safe agent framework https://github.com/pydantic/pydantic-ai
claude-deep-research-skill — 8-phase research with auto-continuation https://github.com/199-biotechnologies/claude-deep-research-skill
MCP & Integrations🔗
MCP Playwright — Browser automation for LLMs https://github.com/executeautomation/mcp-playwright
stealth-browser-mcp — Undetectable browser automation https://github.com/vibheksoni/stealth-browser-mcp
fastmcp — Build MCP servers in minimal Python https://github.com/jlowin/fastmcp
markdownify-mcp — PDFs, images, audio → Markdown https://github.com/zcaceres/markdownify-mcp
MCPHub — Manage multiple MCP servers via HTTP https://github.com/samanhappy/mcphub
Search, Data & LLM Tools 🔍
CK (BeaconBay) — Search code by meaning, not keywords https://github.com/BeaconBay/ck
ExtractThinker — ORM for document intelligence https://github.com/enoch3712/ExtractThinker
OmniRoute — API proxy for 44+ AI providers https://github.com/diegosouzapw/OmniRoute
dlt — LLM-native data pipelines from 5,000+ sources https://github.com/dlt-hub/dlt
simonw/llm — Lightweight CLI for local and remote LLMs https://github.com/simonw/llm
Portkey-AI/gateway — Route requests to 250+ LLMs https://github.com/Portkey-AI/gateway
lmnr — Trace and evaluate agent behavior https://github.com/lmnr-ai/lmnr
Video & More 📽️
LTX-Desktop (Lightricks) — Generate and edit videos locally https://github.com/Lightricks/LTX-Desktop
MetaClaw — Evolve AI agents without GPU https://github.com/aiming-lab/MetaClaw
Vane — AI answering engine with local LLMs https://github.com/ItzCrazyKns/Vane
Where to Find More Skills
skillsmp.com — 80k+ skills, largest catalog
aitmpl.com/skills — Templates, one-command install
skillhub.club — 31k+ skills, AI-rated agentskills.io — Official spec
Install any skill: Personal: ~/.claude/skills/Project: .claude/skills/Clone, copy, restart. Done.
TL;DR
Skills = teach Claude HOW to do things better MCP = give Claude ACCESS to external tools and data GitHub repos = the open-source engines powering it all
Combine all three → unstoppable AI workflow.
That's it. Now go build something.
This list took me hours to compile — scanning 1,000+ repos, testing skills, reading docs. If it saved you time, you know what to do.
I post stuff like this regularly — AI tools, workflows, prompts, and things I actually use. No fluff, no hype, just what works.
Follow me so you don't miss the next one:
Telegram → https://t.me/zodchixquant
If you found a tool I missed — DM me. I'll add it to the next update.