AI agents: what’s taking off on GitHub
The fastest-growing repositories in “AI agents”: plain-language summaries, use cases and a check on star quality. Updated every 30 minutes.
-
6.8
melgarafael/DeskcommCRM
Self-hosted CRM with native AI agents for selling over WhatsApp: agents reply, qualify leads and close sales while your data stays on your server. An open alternative to Kommo, Octadesk and Intercom.
-
6.2
kruzovic7/ai-data-extractor
Python tool that extracts local chat histories from AI coding assistants (Claude Code, Cursor, Windsurf, Aider, Cline/Roo Code and more) into a single normalized JSONL format.
-
6.2
ayghri/i-have-adhd
A skill for a coding agent (Claude Code) that makes it lead with the answer and numbered steps, without preamble or filler. Helps you get concrete actions instead of long-winded reasoning.
-
6.1
tech-leads-club/agent-skills
A registry of validated skills (instruction plugins) for AI coding agents: extends Claude Code, Cursor, Copilot and others with secure, tested capabilities.
-
6.1
jordan-gibbs/hyperresearch
A deep research harness for Claude Code: turns one prompt into a report with verified citations and a persistent source vault. Every source lands in a searchable store reused by later sessions.
-
6.0
NVlabs/SoL-Pi
An extension for Pi that cuts token usage and redundant turns of coding agents via four mechanisms: action fusion, observation packing, log reduction, and context compaction.
-
5.4
alphaXiv/OpenResearch
A local-first workspace for research agents that turns Claude Code, Codex, OpenCode, or Cursor into agents reviewing literature, forming hypotheses, and running experiments. Each direction gets an isolated git worktree…
-
5.4
vastsa/PI-Desktop
Local-first desktop workspace for AI coding agents: opens local projects, connects any model, and extends via plugins, MCP, and Skills. Sessions and files stay on your machine, with privileged actions gated by a…
-
5.4
jihe520/MathModelAgent
An agent for automating mathematical modeling: it analyzes the problem, writes code, fixes errors, and generates a submission-ready Typst paper. Ships as SKILLS for Claude Code or Codex, with a desktop build.
-
5.1
mksglu/context-mode
An MCP server for AI coding agents that sandboxes tool output (up to 98% context reduction), persists session memory in SQLite with FTS5 search, and enforces routing across 17 platforms.
-
5.1
jtydhr88/screenwriting-skills
A set of 24 agent skills (SKILL.md, agentskills.io standard) for Claude Code and OpenAI Codex covering screenwriting, TV writing and dramaturgy, distilled from 45 craft books and 23 volumes of scripts and plays.
-
4.8
nftechie/stonkfly
A fly-connectome simulation (166,700 neurons, 25.6M connections) that feeds Coinbase price charts as visual stimuli and places spot orders via Coinbase AgentKit. Profitable learning has not been demonstrated.
-
4.8
thesysdev/openui
OpenUI is an open standard and full-stack framework for generative UI: an LLM generates interfaces in the compact streaming OpenUI Lang, rendered in React, Vue, or Svelte. Aimed at developers of AI assistants and…
-
4.8
Qiuner/birdview
A tool for coding agents that builds an interactive HTML architecture map before edits and shows which modules the agent plans to touch, letting teams see changes before they happen.
-
4.7
ColeMurray/background-agents
An open-source background coding agent system: the agent works on tasks in isolated sandboxes with a full dev environment, triggered from a web UI, Slack, GitHub, Linear, or webhooks.
-
4.7
TauricResearch/TradingAgents
A framework that simulates a trading firm with LLM agents (analysts, researchers, trader, risk manager) debating markets and producing trading decisions. Built for research, not real trading.
-
4.6
deepseek-ai/deepseek-harness
Open-source agent harness by DeepSeek AI written in TypeScript, built on an everything-is-a-plugin architecture powered by Cordis. It runs as a web UI via npx and serves as a base for composing AI agents from plugins.
-
4.5
rizqinrr/viserys-agent
A self-contained pack of Markdown workflows, reviewer personas, and validation scripts that gives AI coding agents a consistent process across the full development lifecycle.
-
4.5
SpaceDudem/text-humanizer
A Python tool that rewrites AI-generated text through an LLM and translation pipeline to bypass detectors like Turnitin and GPTZero. Supports 8 languages while preserving the original meaning.
-
4.4
petergyang/no-ai-slop
A skill for coding agents (ChatGPT, Claude Code, Codex) that detects and removes 20+ AI slop patterns from writing while preserving the author's voice. It can also flag such patterns and generate them for satire.
-
4.4
Shubhamsaboo/awesome-llm-apps
A collection of 100+ open-source AI agents, agent skills, and RAG apps in Python with ready templates and step-by-step tutorials. Aimed at developers who want to quickly run or adapt an agent for their tasks.
-
4.4
blixvip/MotionClone
MotionClone turns a reference video link or upload into an editable motion-graphics project (HyperFrames) using Codex + ChatGPT, with text, shapes, and animation you can adjust and export to MP4. A local Windows app…
-
4.4
DietrichGebert/ponytail
A skill for AI coding agents (Claude Code, Codex, Copilot CLI, OpenCode) that enforces writing only the minimum necessary code via a laziness ladder: YAGNI, reuse, stdlib, native features. Benchmarks show ~54% less…
-
4.3
linguo2625469/workbuddy2api-panel
Self-hosted OpenAI-compatible gateway that turns Tencent CodeBuddy accounts into a unified /v1/chat/completions API with account pooling, rotation, cooldowns and a web admin panel.
-
4.3
totec448-spec/chat-on-steroids
Cross-platform local app with a Chrome extension that gives ChatGPT access to files, terminal and desktop via MCP. Manages long tasks and splits work across multiple workers.
-
4.2
Tencent/WeKnora
Open-source LLM knowledge platform: turns raw documents into queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.
-
4.2
microsoft/AI-Engineering-Coach
A VS Code extension and GitHub Copilot canvas that analyzes local AI assistant session logs and turns them into a dashboard with metrics, anti-patterns, and context health scoring. No data leaves your machine.
-
4.2
affaan-m/ECC
A performance optimization system for coding agents: a set of agents, skills, memory, and rules that adds a plan-test-review-remember workflow to Claude Code, Codex, Cursor, and other harnesses.
-
4.2
bestagentkits/design-studio-ai
Open-source cloud design workspace for web interfaces, slides, reports, 3D scenes and videos, where humans and AI agents edit the same versioned document via chat, editor, REST, MCP and CLI.
-
4.1
stablyai/orca
Orca is an ADE for running a fleet of parallel coding agents (Claude Code, Codex, OpenCode, etc.) in isolated git worktrees with a unified interface, mobile companion, and remote runtime.
-
4.1
mattpocock/skills
A set of small, composable skills for coding agents like Claude Code and Codex: they help align on a task before coding, build a shared project vocabulary, and follow a red-green-refactor loop.
-
4.0
EvoMap/AutoResearch
An open-source agent workflow for AI/ML research: from idea to a paper-ready evidence package. It runs through experiment planning, implementation, review, execution, and analysis with state persisted to disk.
-
4.0
wechaty/wechaty
Wechaty is a conversational RPA SDK for building chatbots with a unified interface to WhatsApp, WeChat, and other messaging platforms. It lets you write bot code once and run it across platforms in JavaScript, Python,…
-
4.0
github/spec-kit
An open-source toolkit for Spec-Driven Development that turns specifications into executable artifacts and guides the process from requirements to implementation with any AI coding agent.
-
3.9
obra/superpowers
A software development methodology for coding agents: a set of composable skills that make the agent clarify the spec, agree on a plan, and work through subagents with TDD.
-
3.9
Human-Agent-Society/reef
Continual learning infrastructure for self-improving agents: connects agent inference, feedback collection, learning, and versioned delivery of updates. Lets you train model weights or improve an agent's harness…
-
3.9
elder-plinius/FRV1T
Experimental repository containing an encoded 🫐.txt file and a creative prompt for AI agents, offering an "unforbidden fruit" in the form of a blueberry experience. Its purpose and contents are deliberately obfuscated.
-
3.7
dataelement/dsh-desktop
Cross-platform desktop app for running DeepSeek Harness locally: manages agent profiles, plugins, workspaces and sessions, supports third-party models and PPTX generation.
-
3.7
Colinjqq/content-forecast
An AI skill for creators: builds a map of your knowledge and audience, helps pick topics, reviews script drafts, and forecasts views based on past posts.
-
3.7
kunchenguid/firstmate
firstmate is an "agent distro": a portable set of instructions, skills, and scripts that turns a general-purpose coding agent into a "first mate" that spawns and supervises a crew of autonomous agents in separate git…
-
3.7
achimala/dream-loop
An agent skill that builds a game, app, or 3D scene in a loop: generate a target screenshot, build, have a separate critic agent compare and give feedback, repeat until satisfied.
-
3.6
wyuzhi/qisi-video-remix
Agent Skill for creative video remixing: takes a reference video and produces a new story, full script, storyboard, and ready-to-copy per-segment generation prompts.
-
3.5
blader/humanizer
Agent skill that rewrites AI-sounding text to remove telltale patterns while preserving meaning. Ships as a Markdown skill for Claude Code, Codex, Cursor and other agents.
-
3.5
openai/plugins
A curated collection of Codex plugin examples from OpenAI: each plugin ships a plugin.json manifest plus optional skills, MCP configs, agents, and commands. It serves as a reference for building your own plugins and…
-
3.5
tashfeenahmed/freellmapi
Aggregates free tiers from 34 LLM providers (635 endpoints) behind a single OpenAI-compatible /v1 API with smart routing, automatic failover, and encrypted keys. Intended for personal experimentation and use with…
-
3.5
Xu123-Bob/Baize
Open-source AI coding agent CLI with multiple backends (DeepSeek, OpenAI-compatible, local Ollama), tool calling, skills, subagents and a security sandbox.
-
3.4
donvito/codex-astra-luna-orchestrator
A set of Codex configurations where Astra acts as orchestrator and reviewer while Luna runs execution subagents. The installer copies Pro/Plus profiles into the target repo's .codex and .agents.
-
3.4
Git-Agni/prod-FARM-IOS-Core
A self-hosted system for running a farm of real iPhones from a Mac: device registration, WDA/Appium supervision, live video and remote input, a PostgreSQL-backed scheduler, and a built-in TikTok automation plugin.
-
3.4
NxcoreAI/NxMem
A context and long-term memory engine for LLM agents: turns conversations and documents into structured facts, manages short- and long-term memory, and assembles cited context packs.
-
3.4
virgiliojr94/book-to-skill
Converts technical books and other structured documents into agent skills: extracts chapters, glossary, patterns, and cheatsheets that the agent loads on demand. Useful for making a book or docs part of your workflow…
-
3.3
coreyhaines31/marketingskills
A collection of markdown skills that give AI coding agents (Claude Code, Codex, Cursor, Windsurf) ready frameworks for marketing tasks: CRO, copywriting, SEO, analytics, ads, and growth engineering.
-
3.3
Vincentwei1021/video-shotcraft
An AI agent skill for Claude Code and Codex that turns them into a motion-design studio: storyboards, animates, and sound-designs cinematic product promo videos with Remotion using 157 shot recipe cards.
-
3.3
trailhq/Graft
Graft builds a knowledge graph of your codebase as linked markdown files and wires it into Claude Code, Cursor, Codex and other coding agents via MCP so they spend less time re-exploring the repo.
-
3.3
davila7/claude-code-templates
CLI tool for installing and configuring ready-made Claude Code setups: agents, slash commands, MCP integrations, hooks and settings, plus session monitoring. Aimed at developers using Claude Code.
-
3.3
ZJU-REAL/Easel
An open-source AI agent for social media: it discovers trends, creates content, publishes it to Xiaohongshu, Douyin, Zhihu, Bilibili and more, and analyzes what performs. Aimed at content creators and social media…
-
3.3
volcengine/OpenViking
Open-source context database for AI agents: stores knowledge, user memory and skills as a viking:// virtual filesystem with search and tiered context loading.
-
3.2
tigerless-labs/agent-memory
Local long-term memory runtime for AI agents: Markdown files as the source of truth, ranked retrieval over a SQLite index, and sleep-time consolidation. One shared store for Claude Code and Codex, no API keys.
-
3.2
bpy-dev/blender-mcp
MCP server for Blender that gives LLM agents access to Blender's Python API, live scenes, and saved .blend files via headless backends. A fork of Blender Lab MCP with runtime API lookup and benchmark tooling.
-
3.2
agent-launch/agent-launcher
A local desktop app for managing existing coding-agent CLIs: it detects them on your system, configures providers, and runs them from one window.
-
3.1
TencentCloud/TencentDB-Agent-Memory
A team-level memory hub for AI agents that turns conversations, docs, and code into reusable assets (Chat Memory, Skill, Wiki, CodeGraph) with access control and sharing across agents.
-
3.1
maiphucgiang/codebuddy2api
A proxy server that turns a WorkBuddy / CodeBuddy (Tencent) subscription into local OpenAI- and Anthropic-compatible APIs, with tool calling and streaming. Meant for personal use by subscription owners who want to plug…
-
3.1
earthtojake/text-to-cad
A library of agent skills that generate, inspect, and prepare CAD models, drawings, and robot description files (URDF, SRDF, SDF) from text requests.
-
3.0
krmisystems/fantasy-football-manager
MCP server for managing an ESPN fantasy football team: roster sync, legal lineup calculation from projections, and a draft assistant with Monte Carlo simulations. Works with Codex and other MCP clients.
-
2.9
godot-pty/gpty
A Godot- and Rust-based multi-PTY terminal emulator with a tiling pane grid, JSON-RPC/MCP control surface, and a Concept capture engine. It lets AI agents and scripts drive terminals without scraping a TUI.
-
2.9
miuuyy/codex-chatgpt-web
A bridge that lets you use ChatGPT Web (including Pro) as a native model in Codex — with context, tools, streaming and images, without spending Codex quota.
-
2.8
openai/skills
Skills catalog for Codex: folders of instructions, scripts, and resources that AI agents can discover and use to perform tasks repeatably. The repository is marked deprecated; current examples moved to the OpenAI…
-
2.8
xzf-thu/VoiceMem
A memory system for voice agents built on a streaming dual-brain architecture: the left brain stores facts, the right brain handles emotions and persona, with streaming retrieval while the user is still speaking.
-
2.8
gnipbao/gpt6-prompt-writer
A Codex skill for writing, refining, compressing and diagnosing GPT-6 Astra prompts based on official OpenAI guidance; it also audits Skill descriptions and AGENTS.md.
-
2.8
lidge-jun/opencodex
Local proxy server that translates Codex's Responses API into other LLM protocols (Claude, Gemini, Grok, DeepSeek, Ollama, etc.), supporting streaming, tool calls, reasoning tokens and images. It also manages a ChatGPT…
-
2.7
showlab/Show-Harness
Show-Harness is an embodied harness that lets a vision-language model control a robot via discrete action units; aimed at robotics and embodied AI researchers.
-
2.7
OtoDock/oto-dock
Self-hosted AI agent platform: agents built on Claude Code and Codex run on your own Anthropic/OpenAI subscriptions or local models, organized into departments with shared workspaces and sandboxing.
-
2.6
Jakeschincariol/linkedin-agent-skill
A set of eleven free Claude skills for running a LinkedIn account: post generation from 21 hook formulas, comments, replies, profile scoring, a weekly plan, and a humanizer that strips AI fingerprints from drafts.
-
2.6
semantica-agi/semantica
Semantica is graph-native infrastructure for context and accountable AI systems: it turns fragmented enterprise data into a Context Graph and knowledge graph with ontologies, deterministic reasoning, and decision…
-
2.6
ChromeDevTools/chrome-devtools-mcp
An MCP server from Chrome DevTools that lets coding agents control a live Chrome browser: record performance traces, inspect network requests, take screenshots and read console messages. A CLI is also provided for use…
-
2.6
experientiallabs/experiential
Open-source gateway and router for agent LLM workflows: one OpenAI-compatible API over hosted, BYOK and local models with access and budget control, plus training a custom router or model from your traffic.
-
2.6
CopilotKit/OpenBot
A platform for running AI coworkers inside your own infrastructure: each gets a browser, files and tools, with every action going through a gateway for approval and audit. Works with any AG-UI agent.
-
2.6
ApodexAI/FrontierAgent
Open-source agent runtime with a terminal TUI: ReAct and Agent Team modes for long-horizon research and file work in a sandbox, plus an evaluation suite.
-
2.6
viettranx/3dviz-pro-max
An agent skill for Claude Code and Codex that turns an idea into an interactive Three.js or Blender 3D scene, with 223 recipes, 440 knowledge records, 22 kits and 37 runnable studies.
-
2.5
larashero3-dotcom/lieflat-gongwen
A skill for AI agents that turns Chinese official-document (gongwen) style into measurable parameters derived from 1.02M characters of real corpora, helping reproduce 7 document genres.
-
2.5
yc-software/qm
A multiplayer agent harness: each employee gets an isolated workspace with memory, files, and a sandbox, and can also collaborate with the agent in Slack channels and a web app.
-
2.5
chaitanyagiri/munder-difflin
A desktop multi-agent harness that wraps existing terminal coding CLIs (Claude Code, Codex, Gemini, Grok, etc.) into a coordinated office of agents with shared memory, a mailbox, and a clone orchestrator.
-
2.5
vercel-labs/skills
CLI for the open agent skills ecosystem: installs, finds and runs reusable skills from Git repositories across Claude Code, Codex, Cursor, OpenCode and 75+ other agents.
-
2.5
jakubkrehel/skills
A set of AI agent skills for interface design: typography, color, accessibility, layout and UI copy. Installed via npx or as a Claude Code plugin.
-
2.5
trycompai/crm
Open-source TypeScript CRM built around an autonomous AI agent: it runs its own work queue, researches people and companies, and only records facts backed by observed evidence.
-
2.5
punkpeye/awesome-remote-mcp-servers
A curated list of remote MCP servers — hosted endpoints you connect to over a URL with no local install. Helps find ready-made MCP integrations for AI clients.
-
2.5
humanlayer/skills
A collection of Claude Code skills from HumanLayer, installed via npx skills add and invoked as slash commands in a project. Includes improving CLAUDE.md, narrowing React prop types, building agentic loops, and…
-
2.4
MDX-Tom/gpt-instruct
A prompt and test pack for Codex (GPT) that alters model behavior on complex tasks: improves first-pass execution, process continuity and artifact verification. Includes a deploy/rollback script for Codex config.
-
2.4
HammingDev/haiming-app-monetization
A skill for coding agents that assesses app monetization: onboarding, paywalls, subscription options and competitor research. By default it only analyzes the project without changing code.
-
2.3
anthropics/claude-plugins-community
Community plugin marketplace for Claude Cowork and Claude Code: a read-only mirror listing plugins that passed automated security scanning and were approved for distribution. Synced nightly from Anthropic's review…
-
2.3
tradesdontlie/tradingview-mcp
Bridges Claude Code with your local TradingView Desktop via Chrome DevTools Protocol: the AI assistant reads and controls charts and develops Pine Script. Runs fully locally and requires a TradingView subscription.
-
2.3
andrewyng/openworker
An open-source desktop AI coworker that delivers finished work — from security code review to documents and Slack replies. Runs locally with your own model keys and 25+ integrations, with every action approval-gated…
-
2.3
XiaoDuoYa/codex-with-chatgpt
A bridge between the ChatGPT web app and the Codex coding agent: ChatGPT plans and reviews over an OAuth-protected read-only MCP connection while Codex executes. Uses your paid ChatGPT subscription instead of API…
-
2.3
apache/maka
Apache Maka (Incubating) is an agent workspace that keeps a complete append-only RuntimeEvent log of every model message, tool call and permission decision. It ships Desktop, TUI and CLI clients, and you bring your own…
-
2.3
0xranx/OpenContext
A personal context and knowledge store for AI assistants and coding agents: reuses your existing CLI (Codex/Claude/OpenCode) and adds a GUI, MCP server, and Skills to capture, search, and reuse project knowledge across…
-
2.3
tinyhumansai/openhuman
OpenHuman is a local-first agent harness in Rust with persistent memory, agent orchestration, and visual workflows. It pulls data from connected services into a Markdown memory tree and runs agent fleets on durable…
-
2.3
charmbracelet/crush
Terminal-based agentic coding tool that wires your tools, code, and workflows into the LLM of your choice. Supports multiple models, sessions, LSP context, and MCP extensions.
-
2.2
akitaonrails/ai-memory
Long-term memory server for AI coding agents: stores notes as a markdown wiki and hands off context between different agents and machines. Lets you resume a task in another CLI without re-explaining architecture or…
-
2.2
sapientinc/PRAXIST
Autonomous research system for measurable, computer-executable research: coordinates parallel research peers, task-owned evaluation, and durable evidence across generations. Runs on top of an already runnable project…
-
2.2
hardbeat920/monocode
A desktop GUI for running and managing CLI coding agents (Claude Code, Codex, Cursor, Grok Build, OpenCode, and others) using your existing subscriptions. Each tab is a separate agent session.
-
2.2
zhu1090093659/dsh-web
An open-source plugin and theme collection for the DeepSeek Harness (DSH) web GUI: a task kanban board with cron-scheduled agent runs, mobile remote control, an SSH panel, usage statistics, a Git graph and custom…