Coding agents, skills, starters and the stack you ship on. Paid never moves a rank.
Best of Vibe Coding is a ranked list of 1,478 open-source projects for building software with AI: starter templates and SaaS boilerplates, coding agents and assistants, editors, skills and plugins, MCP servers for developers, code review, design to code, and the backends and hosting you ship on. Each entry has strengths, weaknesses and a score from public GitHub signals. Updated 2026-11-19 (UTC).
Tip
π Every project fact-checked on GitHub before it is listed Β Β·Β βοΈ Every entry has strengths and weaknesses; starters say what is wired in Β Β·Β
Note
π€ Independent. We sell none of the tools listed here, and nobody can pay to move up. Ranks come from public GitHub data and are re-checked twice a day; hover a score to see what it is made of.
Looking for AI apps, agents, models and the tools around them? π Best of Open-Source AI
π The same ranking as a website: archestack.com, with a page per project and every score broken down.
1478 projects ranked in 13 categories Β· data from 2026-11-18 Β· re-ranked twice a day
How to read it: rank π₯π₯π₯ then #β 4, #β 5β¦ (equal scores: more stars first) Β· β²βΌ places moved in the last month, π new this month Β· score out of 100 sets the order: hover it to see what it is made of, or read how we rank Β· click a project name for its review Β· β opens another site (Cmd/Ctrl-click for a new tab)
| Rank | Move | Project | Score | Category | Stars |
|---|---|---|---|---|---|
| π₯ | β | deepseek-harness Plugin-based agent harness from DeepSeek with a local Web UI TypeScript Β· MIT Β· π Docs β |
98 | π§© Agent and MCP starters | 284k |
| π₯ | β²10 | claude-mem Persistent cross-session memory plugin for Claude Code and other coding agents π³ Docker Β· Apache-2.0 Β· π Docs β Β· π Site β |
97 | π§© Skills, plugins and prompts | 118k |
| π₯ | β²1 | impeccable Design skill with 24 slash commands for AI coding agents' frontend output Apache-2.0 Β· |
97 | π§© Skills, plugins and prompts | 113k |
| #β 4 | β²3 | codegraph Local code knowledge graph that coding agents query over MCP MIT Β· π Docs β Β· π Site β |
97 | π MCP servers for developers | 78k |
| #β 5 | β²3 | gitnexus Code knowledge graph exposed to coding agents through MCP tools π³ Docker Β· custom license Β· GPU optional Β· |
97 | π MCP servers for developers | 49k |
| #β 6 | βΌ4 | claw-code Rust CLI coding agent harness using Anthropic and OpenAI-compatible APIs π³ Docker Β· MIT Β· π Docs β |
96 | π€ Coding agents and AI editors | 195k |
| #β 7 | β²50 | gstack Markdown slash-command skills that give Claude Code planning, review, QA and ship roles MIT |
96 | π§© Skills, plugins and prompts | 142k |
| #β 8 | β²19 | ruflo Agent harness adding swarms, memory and hooks to Claude Code and Codex MIT Β· |
96 | π§© Skills, plugins and prompts | 82k |
| #β 9 | β²29 | t3code Desktop, web and mobile front end for controlling local coding agent CLIs MIT Β· |
96 | π€ Coding agents and AI editors | 65k |
| #β 10 | π | twenty Open-source CRM extensible with objects, views and agents defined as code TypeScript Β· custom license Β· π Docs β Β· π Site β |
96 | π³ SaaS and app starters | 61k |
Ordered by score: adoption, freshness, how easy it is to run, and a few more signals as we measure them (how we rank).
Most GitHub stars gained from November 11 to 17, 2026 (UTC), from GitHub's daily star counts Β· top 10, this month and an RSS feed: archestack.com/rising
| Project | Stars this week | Rank |
|---|---|---|
| modern-web-guidance | +18,805 (+472%) | #β 65 in Skills, plugins and prompts |
| craft-anything | +8,916 (+32%) | #β 9 in Coding agents and AI editors |
| t3code | +8,354 (+15%) | π₯ in Coding agents and AI editors |
| modern-web-guidance-src | +7,202 (+450%) | #β 122 in Skills, plugins and prompts |
| orca | +6,674 (+5.5%) | #β 16 in Coding agents and AI editors |
- π¬ Ship a chat app this weekend β Chat apps
- π RAG over my data, with citations β RAG and search
- π§© An agent backend I can extend β Agent and MCP starters
- ποΈ A voice agent that talks back β Voice and realtime
- π³ SaaS with auth and billing, ready for AI β SaaS and app starters
- π€ An agent that writes the code for me β Coding agents and AI editors
- π§© Make Claude Code or Cursor better β Skills, plugins and prompts
- π Give my coding agent docs, a browser, GitHub β MCP servers for developers
- π¨ Turn a screenshot into a working page β Coding agents and AI editors
- ποΈ Auth, database and storage in one β Backends, databases and hosting
- π Deploy it on my own server β Backends, databases and hosting
The top open-source picks by score in the matching categories. β means the project's own README calls it an alternative to that product.
| Instead of | Try these open-source picks |
|---|---|
| Cursor | magic β Β· claw-code Β· t3code |
| GitHub Copilot | Tabby β Β· proxyai β Β· cline |
| Claude Code | pydantic-deepagents β Β· claw-code Β· t3code |
| Devin | claw-code Β· t3code Β· codewhale |
| Tabnine | cline Β· codecompanion.nvim Β· kilocode |
| Lovable | windmill β Β· magic β Β· dyad |
| v0 | Onlook β Β· dyad Β· jeecgboot |
| Bolt.new | dyad Β· jeecgboot Β· ai-website-cloner-template |
| Replit | Onlook β Β· magic β Β· dyad |
| CodeRabbit | playwright Β· nx Β· open-code-review |
| Firebase | nhost β Β· trailbase β Β· bun |
| Vercel | coolify β Β· dokploy Β· dokku |
| Heroku | coolify β Β· dokploy Β· dokku |
| Vapi | livekit Β· voice-ui-kit Β· jarvis |
| ShipFast | shipfree β Β· twenty Β· wails |
19 of 325 starters pass all three checks: an env example file at the root (secrets stay out of the code), no .env file committed, and a sign-in library already wired in package.json. A pass is a good start, not a security audit.
| Starter | Sign-in | Category |
|---|---|---|
| hackathon-starter | Passport | π³ SaaS and app starters |
| next-saas-stripe-starter | Better Auth | π³ SaaS and app starters |
| chatbot | Auth.js | π¬ Chat apps |
| velobase-harness | Auth.js | π³ SaaS and app starters |
| start-ui-web | Better Auth | π³ SaaS and app starters |
| personal-agent-template | Better Auth | π§© Agent and MCP starters |
| ack-nestjs-boilerplate | Passport | βοΈ Backend and cloud templates |
| papermark | Auth.js | π³ SaaS and app starters |
| shadcn-admin | Clerk | π³ SaaS and app starters |
| SupabaseAuthWithSSR | Supabase Auth | π RAG and search |
| gemini-chatbot | Auth.js | π¬ Chat apps |
| fulling | Better Auth | π³ SaaS and app starters |
| shipfree | Better Auth | π³ SaaS and app starters |
| saas-starter-kit | Auth.js | π³ SaaS and app starters |
| nextacular | Auth.js | π³ SaaS and app starters |
π§± Starters
- π¬ Chat apps Β· 14
- π RAG and search Β· 21
- π§© Agent and MCP starters Β· 66
- ποΈ Voice and realtime Β· 18
- π³ SaaS and app starters Β· 102
- ποΈ App builder starters Β· 22
- βοΈ AI editors and workflow canvases Β· 10
- βοΈ Backend and cloud templates Β· 72
π οΈ Code with AI
- π€ Coding agents and AI editors Β· 240
- π§© Skills, plugins and prompts Β· 501
- π MCP servers for developers Β· 142
- π§ͺ Code review and testing Β· 48
π Ship it
- ποΈ Backends, databases and hosting Β· 222
Chat interfaces and single-feature text apps you fork as the base of a conversational product. 14 ranked Β· π all reviews Β· π on archestack.com
claude-quickstarts leads with 71, ahead of chat-js (71) and chatbot (65).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | β | claude-quickstarts Claude API sample apps for agents, computer use and browser automation Python Β· MIT Β· π Docs β |
71 | 21k |
| π₯ | π | chat-js Next.js AI chat app scaffold with auth, branching and sharing TypeScript Β· Apache-2.0 Β· |
71 | 1.2k |
| π₯ | βΌ1 | chatbot Next.js chat template with Auth.js, Postgres history and AI Gateway models TypeScript Β· GitHub template Β· Apache-2.0 Β· π safe defaults Β· |
65 | 21k |
| #β 4 | βΌ1 | langchain-nextjs-template Next.js routes for LangChain.js chat, agents, structured output and RAG TypeScript Β· GitHub template Β· MIT Β· |
58 | 2.5k |
| #β 5 | π | boardui React component library for agent UIs, installed as source via CLI TypeScript Β· MIT Β· π Docs β Β· π Site β |
58 | 1.2k |
| #β 6 | βΌ1 | twitterbio Single-form Next.js text generator streaming from Together AI TypeScript Β· MIT Β· |
50 | 1.8k |
| #β 7 | β | gemini-chatbot Next.js chatbot template defaulting to Gemini with NextAuth and Postgres TypeScript Β· GitHub template Β· Apache-2.0 Β· π safe defaults Β· |
45 | 1.4k |
| #β 8 | βΌ4 | chat Nuxt chat app template with GitHub auth, SQLite history and AI SDK streaming Vue Β· GitHub template Β· MIT Β· |
41 | 380 |
| #β 9 | βΌ3 | openai-chatkit-advanced-samples ChatKit feature demos with FastAPI backends and React frontends MIT |
32 | 671 |
| #β 10 | βΌ2 | ai-chat Laravel 12 chat starter streaming replies through Prism to eight providers PHP Β· GitHub template Β· MIT |
26 | 382 |
Top 10 of 14. See all 14, ranked, with reviews β
π‘ How to choose
- Pick by framework first (Next.js, Nuxt, Laravel); porting a chat UI across frameworks costs more than any feature gap.
- Check what is persisted (chat history, files, users) and which database it assumes before you commit.
- Prefer starters that route models through one provider layer so you can swap vendors later.
Retrieval over your own documents or data, answer engines, and natural-language-to-SQL starters. 21 ranked Β· π all reviews Β· π on archestack.com
xerj leads with 87, ahead of sonic (82) and xberg (80).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | β | xerj Elasticsearch-compatible search server that auto-indexes folders for coding agents Rust Β· Apache-2.0 Β· |
87 | 19k |
| π₯ | π | sonic Lightweight schema-less search index that returns object IDs, written in Rust Rust Β· π³ Docker Β· MPL-2.0 Β· |
82 | 21k |
| π₯ | π | xberg Rust document extraction engine with OCR, tables and 15 language bindings Rust Β· π³ Docker Β· MIT Β· |
80 | 9.6k |
| #β 4 | π | qmd Local hybrid search over markdown notes and docs, with CLI and MCP server TypeScript Β· MIT |
78 | 31k |
| #β 5 | βΌ3 | llm-app YAML-configured RAG server that keeps a vector database synced with a document folder Python Β· MIT Β· π Docs β |
77 | 59k |
| #β 6 | βΌ3 | azure-search-openai-demo RAG chat app over your documents using Azure OpenAI and Azure AI Search Python Β· MIT Β· π Docs β |
76 | 7.8k |
| #β 7 | π | manticoresearch C++ search database with full-text, vector and hybrid search over SQL and HTTP C++ Β· π³ Docker Β· GPL-3.0 Β· |
75 | 12k |
| #β 8 | π | paradedb Postgres extension adding BM25 full-text, vector and hybrid search Rust Β· π³ Docker Β· AGPL-3.0 Β· π Docs β Β· π Site β |
75 | 9.5k |
| #β 9 | π | auto-ontology Ontology layer that grounds text-to-SQL agents in certified business definitions Python Β· π³ Docker Β· Apache-2.0 |
73 | 14k |
| #β 10 | βΌ6 | chat-langchain LangChain docs Q&A agent with a Next.js chat frontend TypeScript Β· MIT Β· π Docs β |
56 | 6.5k |
Top 10 of 21. See all 21, ranked, with reviews β
π‘ How to choose
- Match the vector store to what you already run (pgvector, Azure AI Search, a hosted index).
- Look at how ingestion works; a good chat UI with no re-indexing story will stall in production.
- Answer engines need search API keys; count those costs before picking one.
Agent backends and UIs to fork (LangGraph, ADK, OpenAI Agents SDK, Cloudflare Agents) and templates for MCP servers and chat-host apps. 66 ranked Β· π all reviews Β· π on archestack.com
deepseek-harness leads with 98, ahead of composio (92) and conductor (91).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | β | deepseek-harness Plugin-based agent harness from DeepSeek with a local Web UI TypeScript Β· MIT Β· π Docs β |
98 | 284k |
| π₯ | π | composio TypeScript and Python SDKs that give agents authenticated tools across 1000+ apps TypeScript Β· MIT Β· π Docs β Β· π Site β |
92 | 31k |
| π₯ | π | conductor Durable workflow engine that orchestrates microservices and LLM agents from JSON Java Β· π³ Docker Β· Apache-2.0 Β· π Docs β Β· π Site β |
91 | 32k |
| #β 4 | π | a2ui Protocol and renderers for agents to stream native UI instead of text TypeScript Β· Apache-2.0 Β· |
87 | 18k |
| #β 5 | βΌ3 | mcp-use TypeScript framework for building MCP servers with React Views and an Inspector TypeScript Β· MIT Β· |
87 | 11k |
| #β 6 | βΌ3 | computer Durable Object SQLite filesystem with container, shell and JS execution backends TypeScript Β· MIT |
83 | 9.9k |
| #β 7 | π | agent-autoloop Background scheduler that runs CLI agents on events and asks you only when unsure TypeScript Β· MIT Β· π Docs β |
83 | 5.9k |
| #β 8 | π | mirage Virtual shell that mounts S3, Slack and Gmail as one filesystem for agents TypeScript Β· π³ Docker Β· Apache-2.0 Β· π Docs β Β· π Site β |
83 | 3.8k |
| #β 9 | π | agentteams Coordinates a Manager agent and Worker agents in Matrix rooms Go Β· Apache-2.0 Β· π Docs β |
79 | 6.0k |
| #β 10 | βΌ6 | docker-agent Docker CLI plugin that runs multi-agent teams defined in YAML Go Β· π³ Docker Β· Apache-2.0 Β· π Docs β |
79 | 4.7k |
Top 10 of 66. See all 66, ranked, with reviews β
π‘ How to choose
- Choose the framework you want to live with; templates are thin and the framework is the real dependency.
- Check where state lives (Durable Objects, Postgres, a managed platform) and whether that fits your hosting.
- Favor templates that ship tests or an eval hook so agent changes can be checked.
Voice agents, realtime speech-to-speech apps and their web, phone and native clients. 18 ranked Β· π all reviews Β· π on archestack.com
livekit leads with 84, ahead of voice-ui-kit (60) and jarvis (58).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | π | livekit WebRTC SFU server for realtime audio, video and AI agent rooms Go Β· π³ Docker Β· Apache-2.0 Β· |
84 | 22k |
| π₯ | βΌ1 | voice-ui-kit React components and templates for Pipecat voice agent frontends TypeScript Β· BSD-2-Clause Β· π Docs β |
60 | 419 |
| π₯ | π | jarvis Voice interface that brainstorms specs and drives Claude Code sessions on macOS Python Β· custom license |
58 | 1.0k |
| #β 4 | βΌ2 | pipecat-examples Intermediate and advanced example bots built on the Pipecat voice framework TypeScript Β· π³ Docker Β· BSD-2-Clause Β· π Docs β |
56 | 427 |
| #β 5 | βΌ2 | agent-starter-react Next.js voice assistant frontend for LiveKit Agents TypeScript Β· GitHub template Β· MIT Β· π Docs β |
54 | 969 |
| #β 6 | βΌ1 | agent-starter-python Python voice AI agent template for LiveKit Agents Python Β· GitHub template Β· π³ Docker Β· MIT Β· π Docs β Β· π Site β |
51 | 274 |
| #β 7 | βΌ3 | agent-starter-node Node.js voice AI agent starter for LiveKit Agents TypeScript Β· GitHub template Β· π³ Docker Β· MIT Β· π Docs β Β· π Site β |
50 | 118 |
| #β 8 | βΌ2 | agent-starter-android Kotlin and Jetpack Compose voice assistant client for LiveKit Agents Kotlin Β· GitHub template Β· MIT Β· π Docs β |
42 | 106 |
| #β 9 | β | examples Runnable ElevenLabs SDK examples for speech, music, sound effects and voice agents TypeScript Β· MIT Β· π Docs β Β· π Site β |
39 | 635 |
| #β 10 | βΌ3 | agent-starter-swift SwiftUI voice agent client for iOS, macOS and visionOS on LiveKit Swift Β· GitHub template Β· MIT Β· π Docs β |
39 | 97 |
Top 10 of 18. See all 18, ranked, with reviews β
π‘ How to choose
- Pick the transport first (LiveKit, Pipecat/Daily, provider websockets); clients and agents must match.
- Check turn detection and interruption handling; it decides how natural the agent feels.
- Separate frontend and agent starters usually need to be combined; plan for both.
Product boilerplates with auth, billing and a database wired in and AI built in, plus mobile and browser-extension starters. 102 ranked Β· π all reviews Β· π on archestack.com
twenty leads with 96, ahead of wails (94) and next.js (93).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | π | twenty Open-source CRM extensible with objects, views and agents defined as code TypeScript Β· custom license Β· π Docs β Β· π Site β |
96 | 61k |
| π₯ | π | wails Framework for building desktop apps with a Go backend and web frontend Go Β· MIT Β· π Docs β Β· π Site β |
94 | 37k |
| π₯ | π | next.js React framework for full-stack web apps with Rust-based build tooling JavaScript Β· MIT Β· π Docs β Β· π Site β |
93 | 144k |
| #β 4 | π | heroui React component library on React Aria and Tailwind CSS v4 TypeScript Β· Apache-2.0 Β· π Docs β Β· π Site β |
92 | 31k |
| #β 5 | π | astryx React 19 design system with 150+ components, themes and a CLI for agents TypeScript Β· MIT Β· |
91 | 31k |
| #β 6 | π | reflex Python framework for building full-stack web apps without writing JavaScript Python Β· Apache-2.0 Β· π Docs β Β· π Site β |
91 | 29k |
| #β 7 | π | quasar Vue.js framework for web, mobile, desktop and extension apps from one codebase JavaScript Β· MIT Β· π Docs β Β· π Site β |
90 | 27k |
| #β 8 | π | gentelella Vanilla JS and SCSS admin dashboard template with 58 static pages HTML Β· MIT Β· |
90 | 22k |
| #β 9 | π | adminlte Bootstrap 5.3 admin dashboard template in vanilla JavaScript, no jQuery Astro Β· MIT Β· |
89 | 46k |
| #β 10 | π | tabler Bootstrap 5 admin dashboard UI kit with layouts, components and demo pages Astro Β· π³ Docker Β· MIT Β· |
88 | 42k |
Top 10 of 102. See all 102, ranked, with reviews β
π‘ How to choose
- Verify the AI part is real code (model calls, usage metering), not only editor config files.
- Usage-based billing matters for AI costs; check whether metering is included or left to you.
- Weigh the framework lock-in (Wasp, tRPC, Django) against what your team knows.
Prompt-to-app builders and platforms that run coding agents in sandboxes. 22 ranked Β· π all reviews Β· π on archestack.com
dyad leads with 89, ahead of jeecgboot (88) and ai-website-cloner-template (88).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | β²2 | dyad Desktop app that builds apps from prompts on your own machine TypeScript Β· custom license Β· π Site β |
89 | 22k |
| π₯ | β | jeecgboot Java and Vue3 low-code platform with built-in AI apps and knowledge base Java Β· π³ Docker Β· Apache-2.0 Β· |
88 | 49k |
| π₯ | βΌ2 | ai-website-cloner-template Agent skill that rebuilds a website from a URL as a Next.js app TypeScript Β· GitHub template Β· π³ Docker Β· MIT Β· |
88 | 39k |
| #β 4 | π | wasmer WebAssembly runtime and CLI for running code in sandboxes Rust Β· MIT Β· |
85 | 21k |
| #β 5 | βΌ1 | e2b SDKs for running AI-generated code in isolated cloud sandboxes Python Β· π³ Docker Β· Apache-2.0 Β· π Docs β Β· π Site β |
83 | 15k |
| #β 6 | π | appsmith Low-code builder for internal dashboards and admin panels on your own data TypeScript Β· π³ Docker Β· Apache-2.0 Β· π Docs β Β· π Site β |
82 | 41k |
| #β 7 | π | open-slide React slide deck framework that coding agents write and edit TypeScript Β· MIT Β· π Docs β Β· π Site β |
79 | 9.8k |
| #β 8 | π | lowdefy YAML-config framework for building internal web apps and admin tools JavaScript Β· Apache-2.0 Β· π Docs β Β· π Site β |
73 | 3.0k |
| #β 9 | π | magic Self-hosted backend generator that builds secured APIs and apps from plain English C# Β· π³ Docker Β· MIT Β· π Docs β Β· π Site β |
65 | 1.3k |
| #β 10 | π | runtime Firecracker microVM sandbox backend for running untrusted AI agent code Go Β· Apache-2.0 Β· π Docs β Β· π Site β |
64 | 1.9k |
Top 10 of 22. See all 22, ranked, with reviews β
π‘ How to choose
- Check which sandbox runs generated code (E2B, Vercel Sandbox, Cloudflare) and its pricing.
- Look at how repos and credentials are connected; these apps act on real code.
- Expect to bring several API keys; the templates are platforms, not single-page demos.
Rich-text editors with AI commands and node-based canvases for chaining model calls. 10 ranked Β· π all reviews Β· π on archestack.com
penpot leads with 89, ahead of windmill (89) and next-ai-draw-io (86).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | π | penpot Open-source design and prototyping tool with real-time collaboration and design tokens Clojure Β· π³ Docker Β· MPL-2.0 Β· |
89 | 63k |
| π₯ | π | windmill Workflow engine and app builder where scripts become UIs, APIs and cron jobs Rust Β· π³ Docker Β· custom license Β· |
89 | 19k |
| π₯ | π | next-ai-draw-io Chat with an AI to create and edit draw.io diagrams TypeScript Β· π³ Docker Β· Apache-2.0 Β· |
86 | 37k |
| #β 4 | π | plate Rich-text editor framework for React built on Slate, with plugins and AI TypeScript Β· custom license Β· π Docs β Β· π Site β |
86 | 17k |
| #β 5 | π | fresh Terminal text editor with VS Code-style keybindings, mouse support and LSP Rust Β· GPL-3.0 Β· π Docs β Β· π Site β |
74 | 9.5k |
| #β 6 | π | builder Visual editor and SDKs for drag-and-drop editing with your own components TypeScript Β· MIT Β· π Docs β Β· π Site β |
66 | 8.9k |
| #β 7 | βΌ6 | plate-playground-template Next.js rich-text editor template on Plate with AI commands TypeScript Β· GitHub template Β· MIT Β· π Docs β |
53 | 243 |
| #β 8 | βΌ5 | workflow-builder-template Visual AI workflow builder on Workflow DevKit with real integrations TypeScript Β· GitHub template Β· Apache-2.0 |
47 | 1.2k |
| #β 9 | βΌ7 | editor Nuxt template for a Notion-like TipTap editor with AI completions TypeScript Β· GitHub template Β· MIT Β· |
40 | 171 |
| #β 10 | βΌ6 | tersa Node canvas for chaining text, image and video models via AI Gateway TypeScript Β· MIT |
18 | 1.0k |
π‘ How to choose
- For editors, pick the editor engine you can extend (TipTap, Plate) before the AI features.
- For canvases, check whether workflows persist server-side or only in the browser.
- Collaboration features add infrastructure; skip them if you do not need multi-user editing.
Backend service templates (FastAPI, Express, Hono) that expose models or agents over an API, and vendor reference apps for Azure or Google Cloud. 72 ranked Β· π all reviews Β· π on archestack.com
fastapi leads with 90, ahead of trpc (90) and fiber (90).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | π | fastapi Python web framework for typed REST APIs with automatic OpenAPI docs Python Β· MIT Β· π Docs β Β· π Site β |
90 | 105k |
| π₯ | π | trpc TypeScript library for typesafe APIs without schemas or code generation TypeScript Β· MIT Β· π Docs β Β· π Site β |
90 | 41k |
| π₯ | π | fiber Express-style HTTP web framework for Go, built on fasthttp Go Β· MIT Β· π Docs β Β· π Site β |
90 | 40k |
| #β 4 | π | generator-jhipster Code generator for Java apps with Angular, React or Vue frontends TypeScript Β· π³ Docker Β· Apache-2.0 Β· π Docs β Β· π Site β |
90 | 22k |
| #β 5 | π | serverless CLI that deploys AWS Lambda code and infrastructure from serverless.yml JavaScript Β· custom license Β· π Docs β Β· π Site β |
88 | 47k |
| #β 6 | π | echo Minimalist Go web framework with a radix-tree router and middleware Go Β· MIT Β· π Docs β Β· π Site β |
88 | 33k |
| #β 7 | π | douyin_tiktok_download_api Self-hosted REST and MCP API for Douyin, TikTok and Bilibili data Python Β· π³ Docker Β· Apache-2.0 Β· |
86 | 21k |
| #β 8 | π | vapor HTTP web framework for building APIs and websites in Swift Swift Β· MIT Β· π Docs β |
85 | 26k |
| #β 9 | π | hono TypeScript web framework for Workers, Deno, Bun and Node.js TypeScript Β· MIT Β· π Docs β Β· π Site β |
84 | 33k |
| #β 10 | π | gf Modular Go application framework with an HTTP server and CLI scaffolding Go Β· MIT Β· π Docs β Β· π Site β |
83 | 13k |
Top 10 of 72. See all 72, ranked, with reviews β
π‘ How to choose
- Only pick these if you are already on that cloud; the value is in the infra wiring.
- Check the IaC tool (azd/Bicep, Terraform) and the identity model before forking.
- Budget for managed services these templates provision by default.
Agents, assistants and editors that read your repo, write code and run commands, including tools that turn designs and screenshots into code. 240 ranked Β· π all reviews Β· π on archestack.com
claw-code leads with 96, ahead of t3code (96) and codewhale (94).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | β | claw-code Rust CLI coding agent harness using Anthropic and OpenAI-compatible APIs π³ Docker Β· MIT Β· π Docs β |
96 | 195k |
| π₯ | β²8 | t3code Desktop, web and mobile front end for controlling local coding agent CLIs MIT Β· |
96 | 65k |
| π₯ | β²3 | codewhale Terminal coding agent in Rust that works with 40+ model providers π³ Docker Β· MIT Β· GPU optional Β· π Docs β Β· π Site β |
94 | 41k |
| #β 4 | βΌ1 | oh-my-openagent Terminal coding agent with multi-model routing and git-backed memory custom license Β· π Docs β Β· π Site β |
93 | 71k |
| #β 5 | βΌ1 | goose Rust AI agent with desktop app, CLI and API, extended through MCP π³ Docker Β· Apache-2.0 Β· π Docs β |
93 | 57k |
| #β 6 | β²5 | Archon YAML workflow engine that runs AI coding agents through fixed dev steps π³ Docker Β· MIT Β· π Docs β Β· π Site β |
93 | 24k |
| #β 7 | βΌ5 | pi Minimal terminal coding agent extensible with skills and packages MIT Β· π Docs β Β· π Site β |
92 | 123k |
| #β 8 | βΌ1 | opencode Terminal-based AI coding agent with build and plan modes MIT Β· π Docs β Β· π Site β |
91 | 228k |
| #β 9 | π | craft-anything Rust engine that inspects a website, app or game and has an agent rebuild it Apache-2.0 Β· π Site β |
91 | 37k |
| #β 10 | β²7 | paseo Desktop, mobile and CLI client that runs many coding agents in parallel worktrees π³ Docker Β· custom license Β· π Docs β Β· π Site β |
91 | 24k |
Top 10 of 240. See all 240, ranked, with reviews β
π‘ How to choose
- Check which models it can use: your own keys, local models, or only the vendor's.
- See what it may do without asking (edit files, run shell, push) and how you approve.
- Prefer agents that work in a branch or worktree you can throw away.
Add-ons that make coding agents better (skills, plugins, rules, hooks, memory) and prompt collections to learn from or reuse. 501 ranked Β· π all reviews Β· π on archestack.com
claude-mem leads with 97, ahead of impeccable (97) and gstack (96).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | β²6 | claude-mem Persistent cross-session memory plugin for Claude Code and other coding agents π³ Docker Β· Apache-2.0 Β· π Docs β Β· π Site β |
97 | 118k |
| π₯ | β | impeccable Design skill with 24 slash commands for AI coding agents' frontend output Apache-2.0 Β· |
97 | 113k |
| π₯ | β²23 | gstack Markdown slash-command skills that give Claude Code planning, review, QA and ship roles MIT |
96 | 142k |
| #β 4 | β²8 | ruflo Agent harness adding swarms, memory and hooks to Claude Code and Codex MIT Β· |
96 | 82k |
| #β 5 | βΌ1 | ecc Agents, skills, hooks and memory bundle for Claude Code and other harnesses π³ Docker Β· MIT Β· π Site β |
95 | 306k |
| #β 6 | βΌ5 | caveman Shortens coding-agent replies and compresses tool output before the model reads it π³ Docker Β· Apache-2.0 Β· |
95 | 121k |
| #β 7 | β²1 | oh-my-claudecode Multi-agent orchestration plugin for Claude Code with team pipelines MIT Β· π Docs β Β· π Site β |
95 | 41k |
| #β 8 | βΌ5 | skills Agent skills for planning, TDD, debugging and code review MIT Β· π Site β |
94 | 327k |
| #β 9 | β²1 | agent-skills Markdown engineering skills and slash commands for AI coding agents MIT Β· π Docs β |
93 | 119k |
| #β 10 | β²1 | archify Agent skill that turns descriptions or repos into interactive HTML diagrams Apache-2.0 Β· |
93 | 98k |
π§ Go deeper: Skillselion β indexes skills and plugin marketplaces by installs.
Top 10 of 501. See all 501, ranked, with reviews β
π‘ How to choose
- Check which agent it targets and whether it works with the version you run.
- Read what it adds to every prompt; big rule sets cost tokens on each call.
- Prefer add-ons you can turn off per project.
MCP servers that give coding agents tools: docs lookup, browsers, GitHub, databases and code navigation. 142 ranked Β· π all reviews Β· π on archestack.com
codegraph leads with 97, ahead of gitnexus (97) and playwright-mcp (93).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | β | codegraph Local code knowledge graph that coding agents query over MCP MIT Β· π Docs β Β· π Site β |
97 | 78k |
| π₯ | β | gitnexus Code knowledge graph exposed to coding agents through MCP tools π³ Docker Β· custom license Β· GPU optional Β· |
97 | 49k |
| π₯ | π | playwright-mcp MCP server that lets LLMs drive a browser through Playwright accessibility snapshots π³ Docker Β· Apache-2.0 Β· π Site β |
93 | 40k |
| #β 4 | β²2 | code-review-graph Local Tree-sitter code graph that gives coding assistants review context over MCP π³ Docker Β· MIT Β· π Docs β Β· π Site β |
92 | 33k |
| #β 5 | π | fastmcp Python framework for building MCP servers, clients, and interactive apps Apache-2.0 Β· π Docs β Β· π Site β |
92 | 29k |
| #β 6 | βΌ3 | serena MCP server giving coding agents symbol-level code search, editing and refactoring π³ Docker Β· custom license Β· π Docs β |
91 | 32k |
| #β 7 | β | n8n-mcp MCP server exposing n8n node docs, templates and workflow management to AI assistants π³ Docker Β· MIT Β· |
91 | 23k |
| #β 8 | π | cua Desktop automation tools and VM sandboxes for computer-use agents π³ Docker Β· MIT Β· π Docs β Β· π Site β |
90 | 33k |
| #β 9 | π | inspector Inspector for testing and debugging MCP servers via web, CLI or terminal UI π³ Docker Β· custom license Β· π Site β |
90 | 11k |
| #β 10 | βΌ6 | context7 MCP server and CLI that feed current library docs to coding agents MIT Β· π Docs β Β· π Site β |
89 | 65k |
π§ Go deeper: Glama β grades thousands of MCP servers on security and quality.
Top 10 of 142. See all 142, ranked, with reviews β
π‘ How to choose
- Check what it can change (push, delete, run) and how that is scoped.
- Prefer servers that run locally over ones that send your code to a third party.
- Look at how many tool calls a task takes; chatty servers burn context.
AI tools that review pull requests, write tests or find bugs, run in your CI or on your machine. 48 ranked Β· π all reviews Β· π on archestack.com
playwright leads with 94, ahead of nx (93) and open-code-review (92).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | π | playwright Cross-browser automation and end-to-end test framework for Chromium, Firefox and WebKit Apache-2.0 Β· π Docs β Β· π Site β |
94 | 101k |
| π₯ | π | nx Monorepo build system with task caching, affected-only runs and CI integration MIT Β· π Docs β Β· π Site β |
93 | 29k |
| π₯ | βΌ2 | open-code-review Go CLI that reviews Git diffs with an LLM agent and line-level comments Apache-2.0 Β· π Docs β Β· π Site β |
92 | 54k |
| #β 4 | π | maestro YAML-based UI and end-to-end test framework for Android, iOS and web Apache-2.0 Β· π Docs β Β· π Site β |
89 | 16k |
| #β 5 | π | bend Dependent-typed language where AI-written code must prove rules in LAWS.bend Apache-2.0 Β· GPU optional Β· |
88 | 26k |
| #β 6 | π | midscene Vision-based E2E testing in natural language for web, mobile and desktop MIT Β· π Docs β Β· π Site β |
88 | 16k |
| #β 7 | π | react-doctor CLI that scans React codebases for state, performance, security and accessibility issues custom license Β· π Docs β |
88 | 15k |
| #β 8 | π | knip Finds unused dependencies, exports and files in JavaScript and TypeScript projects ISC Β· π Docs β Β· π Site β |
88 | 13k |
| #β 9 | βΌ7 | pr-agent LLM-based pull request reviewer that runs as a GitHub Action, CLI or webhook π³ Docker Β· MIT Β· π Docs β |
87 | 16k |
| #β 10 | π | infracost Cloud cost estimates for Terraform and CloudFormation in the terminal, editor and PRs π³ Docker Β· Apache-2.0 Β· π Docs β Β· π Site β |
87 | 13k |
Top 10 of 48. See all 48, ranked, with reviews β
π‘ How to choose
- Check where it runs: your CI with your keys, or the vendor's service.
- Try it on a real PR; how much noise it makes matters more than its feature list.
- See which languages and test frameworks it supports.
Open-source backends you ship apps on (auth, database, storage, APIs) and self-hosted platforms that deploy them from Git or Docker. 222 ranked Β· π all reviews Β· π on archestack.com
bun leads with 94, ahead of coolify (93) and appwrite (93).
| Rank | Move | Project | Score | Stars |
|---|---|---|---|---|
| π₯ | π | bun JavaScript and TypeScript runtime with bundler, test runner and package manager π³ Docker Β· custom license Β· π Docs β Β· π Site β |
94 | 97k |
| π₯ | π | coolify Self-hosted Heroku and Vercel alternative that deploys apps and databases over SSH π³ Docker Β· Apache-2.0 Β· π Docs β Β· π Site β |
93 | 66k |
| π₯ | βΌ2 | appwrite Backend with auth, databases, storage, functions and an MCP server for agents π³ Docker Β· BSD-3-Clause Β· |
93 | 58k |
| #β 4 | π | strapi Headless CMS that generates REST and GraphQL APIs from content models π³ Docker Β· custom license Β· |
92 | 74k |
| #β 5 | π | traefik HTTP reverse proxy and load balancer that auto-discovers routes from orchestrators π³ Docker Β· MIT Β· π Docs β Β· π Site β |
92 | 66k |
| #β 6 | π | tidb Distributed MySQL-compatible SQL database with row and columnar storage π³ Docker Β· Apache-2.0 Β· |
92 | 41k |
| #β 7 | π | medusa Node.js commerce platform with modular products, carts, orders and payments custom license Β· π Docs β Β· π Site β |
92 | 37k |
| #β 8 | π | duckdb In-process analytical SQL database that queries CSV, Parquet and JSON files directly MIT Β· π Docs β Β· π Site β |
91 | 43k |
| #β 9 | π | compose CLI plugin that runs multi-container apps from a compose.yaml file π³ Docker Β· Apache-2.0 Β· π Docs β |
91 | 39k |
| #β 10 | π | remix Web framework built from composable Fetch API packages for any JS runtime MIT Β· π Site β |
91 | 33k |
Top 10 of 222. See all 222, ranked, with reviews β
π‘ How to choose
- Check what you get self-hosted versus only in the managed cloud.
- See how it handles auth, row-level security and file storage.
- Look at the client SDKs for your stack.
Every project gets a score out of 100 from the signals below. A signal we cannot measure yet is left out and its weight goes to the others, so nobody loses points for things we have not checked. The score sets the order; π₯π₯π₯ mark the top three places in each table.
| Signal | Weight | What it looks at | Shown as (80+ Β· 50+ Β· 30+ Β· lower) | Status |
|---|---|---|---|---|
| β Adoption | 15% | GitHub stars, log-scaled, ranked half within the category and half across the list | widely used Β· popular Β· known Β· niche | live |
| π Momentum | 10% | stars gained in the last 30 days, ranked within the category | surging Β· rising Β· steady Β· flat | soon |
| π Freshness | 15% | last commit (full marks under 90 days, zero at a year) and last release | active Β· recent Β· slowing Β· quiet | live |
| π§ Maintenance | 15% | issues closed vs opened, release cadence, commits in 90 days | healthy Β· fair Β· patchy Β· weak | live |
| β Verified build | 30% | we start it in CI and wait for a healthy response | passes Β· untested Β· fails | soon |
| π Docs | 5% | install section, environment docs, upgrade notes | thorough Β· good Β· thin Β· missing | soon |
| π€ Agent-ready | 5% | CLAUDE.md, AGENTS.md, skills, llms.txt | ready Β· partly Β· minimal Β· none | live |
| π Privacy | 5% | no telemetry or opt-out, works fully offline | private Β· mostly private Β· some telemetry Β· leaky | soon |
Penalties: no license β10. Coming: data must leave your host β10 Β· unpatched security advisory β20. Archived projects are removed. Paid placements never touch the score.
The weights are our judgment, not science. Think one is wrong? Tell us. Method changes: 2026-10-14: the star-growth anomaly penalty is dropped, since a star chart cannot tell a big launch from bought stars Β· 2026-10-13: Maintenance weighs 15% (was 10%) and Momentum 10% (was 15%), so steady upkeep counts more than a burst of stars Β· 2026-10-12: penalties we do not check yet are marked coming; only "no license" has ever been applied Β· 2026-10-09: Docker alone earns no points, medals mark places 1β3 instead of score bands.
- π§± Starters are templates and reference apps you clone and build on. π οΈ Code with AI is what you build with. π Ship it is the stack you deploy on, AI or not.
- π Found in curated lists, app stores and template galleries, then fact-checked on GitHub: stars, last commit, license, Docker files.
- βοΈ Written from the README, never copied from other lists: what it does, what it needs, strengths and weaknesses as claims you can check. Specs say
unknownrather than guess. - π Kept fresh: entries are rewritten when the README or the latest release changes; projects quiet for 12 months get marked stale, archived ones drop off.
- π« No pay-to-rank. Sponsors and affiliate links, if any, are labeled and never touch the order.
β Add a project Β· π οΈ Fix wrong data Β· πͺ Opt out (done within 24 hours). Everything here is generated, so use the forms instead of editing files.
π Where candidates come from
These lists, app stores and galleries (facts and links only, no text copied), plus community submissions: awesome-llm-apps Β· awesome-saas-boilerplates Β· awesome-opensource-boilerplates Β· awesome-langgraph Β· awesome-langchain Β· awesome-supabase Β· voiceai Β· awesome-nextjs Β· gh-search-ai-templates Β· gh-search-llm-starters Β· gh-search-rag-templates Β· vercel Β· langchain-ai Β· openai Β· anthropics Β· livekit-examples Β· pipecat-ai Β· copilotkit Β· assistant-ui Β· cloudflare Β· run-llama Β· azure-samples Β· aws-samples Β· google-gemini Β· googlecloudplatform Β· supabase-community Β· get-convex Β· awesome-claude-code Β· awesome-claude-skills Β· awesome-mcp-servers-dev Β· gh-search-claude-code Β· gh-search-coding-agent Β· gh-search-vibe-coding Β· gh-search-ai-coding Β· gh-search-agent-skills Β· gh-search-baas Β· gh-search-paas Β· gh-search-boilerplate Β· gh-search-saas-boilerplate Β· gh-search-ai-app-builder Β· gh-search-lovable Β· gh-search-mcp-apps Β· awesome-vibe-coding Β· gh-rising-new Β· gh-rising-half-year Β· gh-trending Β· repodir-recent.
Data (data/, this README, reviews/) is CC BY 4.0; see LICENSE-DATA. Code is MIT; see LICENSE. Project names and descriptions belong to their owners.
Curated with care, powered by public GitHub data, funded by freelance gigs. Found a gem? Tips keep the servers happy, but rankings stay honest. π
For freelance collabs or coffee-fueled conversations, say hello on LinkedIn.
Made with β by Archestack Β· on the web at archestack.com Β· see also Best of Open-Source AI
