Hyperautomation — This month

01

microsoft/autogen — A programming framework for agentic AI

GitHub

github.com 61663 pts 16 Oct

02

DayuanJiang/next-ai-draw-io — Draw and edit draw.io diagrams by chatting with an AI. Web app, desktop app, and an MCP server for Claude Code, Cursor, Codex and VS Code.

AIThe open-source Next AI Draw.io project lets users create and edit diagrams via conversational AI, offering an MCP server to integrate diagramming directly into coding agents.

GitHub

github.com 36548 pts +1/h 16 Oct

03

windmill-labs/windmill — Code-first platform for workflows, apps and AI agents, on the fastest self-hostable orchestrator (benchmarked against Airflow, Prefect and Temporal). Business teams build with AI, engineers control what ships. Open-source alternative to Lovable, Retool and n8n.

GitHub

github.com 18354 pts +1/h 25 Oct

04

n8n-io/self-hosted-ai-starter-kit — The Self-hosted AI Starter Kit is an open-source template that quickly sets up a local AI environment. Curated by n8n, it provides essential tools for creating secure, self-hosted AI workflows.

AIn8n released a Docker Compose starter kit combining n8n, Ollama, Qdrant, and PostgreSQL, letting developers deploy fully local, privacy-compliant AI automation workflows.

GitHub

github.com 15297 pts 12 Oct

05

autotrust/JEV-27B-VL — image-text-to-text

AIThe JEV-27B-VL vision-language model enables fast, single-pass visual decision-making, allowing UI automation and robotic agents to select actions directly from screenshots in roughly 240 milliseconds.

HF

huggingface.co 6978 pts +28/h 5 Oct

06

facebook/sam3 — mask-generation

HF

huggingface.co 5646 pts +6/h 11 Oct

07

baidu/Unlimited-OCR — image-text-to-text

HF

huggingface.co 4865 pts 12 Oct

08

deepseek-ai/DeepSeek-V4.1-Flash-1017 — image-text-to-text

HF

huggingface.co 4341 pts +15/h 17 Oct

09

kajisho5/ffmpeg-skill — Local video editor for coding agents: 42 FFmpeg tools with a machine-readable contract and an MCP server, for Claude Code, Cursor and Codex. Cut, captions, reframe, loudness, platform exports. No cloud, no API keys.

AIA new open-source MCP server and skill package exposes 42 local FFmpeg tools to coding agents, enabling video editing workflows without external APIs or cloud services.

GitHub

github.com 4283 pts +3/h 15 Oct

10

pyannote/speaker-diarization-community-1 — automatic-speech-recognition

HF

huggingface.co 3013 pts +4/h 16 Oct

11

cranebeaker/coop-anything — Co-op Anything: point your AI agent at any single-player game and play it with friends (online and/or splitscreen!). MCP server and playbook to let you reverse engineer, fix, test, ship and share what it learns.

AICo-op Anything provides an MCP server and playbook that enables AI coding agents to reverse engineer, develop, and test multiplayer mods for single-player video games.

GitHub

github.com 1787 pts 21 Oct

12

hujjun/mcd-saver-strategist — 麦当劳省钱点餐助手（麦门精算师）｜麦当劳优惠券组合优化 Skill：拉取麦当劳菜单与全部可用优惠券，算出最省的点餐组合，巨无霸套餐 ¥47 → ¥21.9（省 53%）。McDonalds MCP（麦当劳 MCP）coupon optimizer · McDonalds China MCP Server · Agent Skill for WorkBuddy · 麦当劳程序员创意开发大赛参赛作品

AIAn open-source agent skill connects WorkBuddy to the McDonald's China MCP server, letting developers automate coupon optimization, price calculation, and order placement via programmatic APIs.

GitHub

github.com 1605 pts 16 Oct

13

ollaya-dev/ollaya — Jev-compatible API server for open decision models. Ollama for decision models: pull and serve 24 model families (Laya, pplx-decider, Clef, Winnow, Kev...) locally or self-hosted behind /v1/systemone. TypeSafe SDK works by changing one env var. CLI, daemon, desktop app, Docker, MCP server, public bench.

AIOllaya released an open-source API server to locally serve decision models, allowing developers to run specialized decision-making logic locally via an MCP server or SDK.

GitHub

github.com 1433 pts 23 Oct

14

HexRaysSA/ida-mcp — Official Hex-Rays IDA MCP Server.

AIHex-Rays released an official IDA MCP server, enabling AI agents to automate binary analysis workflows in IDA Pro via stdio or HTTP interfaces.

GitHub

github.com 1397 pts 6 Oct

15

ASCIT31/Dark-Moon — Open-source autonomous AI penetration testing. 50 specialist agents across web, API, cloud, Active Directory, Kubernetes, CI/CD and AI/LLM (OWASP LLM Top 10). Runs on a local LLM behind a Privacy Gateway. Integrates with GitHub, GitLab, Jenkins, VS Code, JetBrains, n8n, Grafana & Splunk.

GitHub

github.com 1183 pts +3/h 23 Oct

16

alexgaoth/awesome-reverse-engineered — Don't reverse engineer it twice. 16,000+ open-source tools built by reverse engineering closed apps, APIs, devices, protocols, file formats, games and AI tools. Search site + agent skill + MCP server, refreshed nightly.

AIAwesome-reverse-engineered cataloged over 16,000 open-source tools with an MCP server and agent skill, helping builders integrate closed APIs, devices, and protocols into automated workflows.

GitHub

github.com 1049 pts +7/h 5d ago

17

bytechefhq/bytechef — Open-source AI agents and workflow automation. Self-host or embed in your SaaS. Apache 2.0 alternative to n8n and Zapier.

AIByteChef provides an Apache 2.0 open-source platform to build, self-host, and embed AI agents and workflow automation directly into SaaS applications.

GitHub

github.com 1043 pts 7 Oct

18

hanzohanzhe/Grandet — Find the cheapest API for any model and connect it from your AI agent: Grandet Skill and an anonymous MCP server, backed by listed prices from 1,800+ official, inference and relay providers on grandet.ai.

AIGrandet released an MCP server and skill that lets AI agents dynamically query model pricing across 1,800 providers to automatically route tasks to the cheapest API.

GitHub

github.com 1037 pts 5d ago

19

LiCinCan/blender-ai-mcp-addon — AI assistant and MCP server for Blender: visual chat, scene tools, multi-agent workflows. Works with OpenAI, ChatGPT, Claude, Gemini, Ollama, OpenCode. Blender AI addon and plugin, copilot for 3D, bpy Python scripting, MCP integration.

AIAn open-source Blender add-on introduces a built-in Model Context Protocol server, enabling external AI agents and LLMs to inspect scenes, run Python scripts, and automate 3D modeling tasks.

GitHub

github.com 976 pts 18 Oct

20

scenario-labs/blender-plugin — Scenario for Blender: AI images, video, 3D and PBR materials in the viewport, plus a local MCP server for agents. Blender 5.0+ extension, GPL-3.0-or-later.

AIA new Blender extension embeds generative AI tools directly into the viewport and includes a local Model Context Protocol server, enabling AI agents to automate 3D workflows.

GitHub

github.com 885 pts 17 Oct

21

windmill 1.824.0

AIWindmill 1.824.0 introduces an AI decision flow step, adds a Notion OAuth provider, and expands Google and Slack integration scopes.

Release

windmill 5 Oct

22

kestra 2.0.5

AIKestra 2.0.5 introduces a centralized HTTP retry mechanism, adds configurable worker job buffers, and fixes several execution retry, loop concurrency, and breakpoint handling issues.

Release

kestra 5 Oct

23

n8n n8n@3.0.0

AIn8n 3.0.0 fixes agent node credential access, strengthens streaming and execution security, and enhances AI assistant and task runner stability across automated workflows.

Release

n8n 5d ago

24

kestra 2.0.8

Release

kestra 5d ago

25

windmill 1.832.0

AIWindmill 1.832.0 tracks Python versions per instance and defaults new ones to Python 3.13, while improving variable storage capacity and S3 proxy security.

Release

windmill 13 Oct

26

kestra 2.0.6

AIKestra 2.0.6 converts the Schedule trigger away from polling and resolves bugs with subflow retries, task duplication, and looped executions.

Release

kestra 13 Oct

27

prefect 3.9.0 - Pendulum's last swing

AIPrefect 3.9.0 drops Python 3.10 support, deprecates Pendulum datetimes, removes legacy runner paths, and improves automation resilience during websocket outages.

Release

prefect 21 Oct

28

trigger.dev .docker.4.8.0: chore: release v4.8.0 (#5044)

AITrigger.dev v4.8.0 adds hosted webhooks and agent chat routing, letting developers automatically verify inbound events, pause runs until webhooks arrive, and connect AI agents to external chat channels.

Release

trigger.dev 23 Oct

29

langgraph langgraph-sdk==0.4.9

AILangGraph SDK 0.4.9 adds a Priority feature for thread settings, allowing developers to manage the execution precedence of concurrent agent workflows.

Release

langgraph 23 Oct

30

langgraph langgraph==1.2.13

AILangGraph 1.2.13 fixes checkpoint replay and branching issues, improving state management and interrupt tracking reliability for complex multi-agent workflows.

Release

langgraph 5 Oct

31

airflow providers-duckdb/0.2.0

Release

airflow 5 Oct

32

mcp servers 1.0.0

AIModel Context Protocol reference servers reached version 1.0.0, providing stable packages for filesystem, memory, and task execution integrations in AI agent workflows.

Release

mcp servers 11 Oct

33

temporal 1.33.0-164.2

Release

temporal 24 Oct

34

airflow providers-vespa/0.1.3

Release

airflow 24 Oct

35

windmill 1.828.0

AIWindmill 1.828.0 adds retries for asynchronous script runs, integrates new AI models into chat and agent steps, and improves log capture for non-UTF-8 outputs.

Release

windmill 8 Oct

36

google/embeddinggemma-2 — feature-extraction

HF

huggingface.co 2994 pts +1/h 6 Oct

37

autotrust/GEV-26B-Decide — text-classification

AIAutoTrust released GEV-26B-Decide, an open-weights decision model that enables sub-100ms computer use and tool selection by combining fast single-pass inference with adaptive reasoning.

HF

huggingface.co 2099 pts +19/h 5 Oct

38

mistralai/Mistral-Large-4-1T-A52B — image-text-to-text

HF

huggingface.co 1454 pts +1/h 3d ago

39

RecursiveSI/flint — text-generation

AIRecursive released Flint, an open-weights 40B mixture-of-experts model optimized for coding agents, multi-turn tool use, and codebase automation with 2.5B active parameters per token.

HF

huggingface.co 1296 pts +1/h 21 Oct

40

openbmb/MiniCPM-V-4.7-35B-A3B — image-text-to-text

HF

huggingface.co 1073 pts +1/h 13 Oct

41

Qwen/Qwen-Image-2.1-Turbo — text-to-image

AIQwen released Qwen-Image-2.1-Turbo, an accelerated 7B text-to-image model using eight denoising steps, enabling faster, lower-latency image generation inside Diffusers-based automation pipelines.

HF

huggingface.co 1022 pts 9 Oct

42

Nanbeige/Nanbeige4.3-33B — text-generation

HF

huggingface.co 975 pts 25 Oct

43

XHToken/Spark-X2.6-4B — text-generation

HF

huggingface.co 948 pts +1/h 22 Oct

44

Aleph-Alpha/Kolibri-1 — text-generation

AIAleph Alpha released Kolibri-1, an open-weight 78B bilingual MoE model with native tool calling and long-context support for German and English agentic automation workflows.

HF

huggingface.co 928 pts 3 Oct

45

LingCore/WinShun — Win顺：免费开源的 Windows 快速搜索启动器。双击 Ctrl，拼音秒搜文件、全文搜索 Word/Excel/PDF、剪贴板历史（代替 Win+V）、对话框搜索、任务栏搜索、插件、AI 助手 MCP。Free Windows search launcher: pinyin file search, full-text search, clipboard history, MCP server. Everything/Listary alternative.

GitHub

github.com 792 pts +7/h 15 Oct

46

ctoldigital/axuse — Accessibility-first browser & computer use for AI agents — one MCP server for every harness (Codex, Claude Code, OpenCode, Antigravity). Instruments the DOM/accessibility tree via CDP instead of screenshot loops: diffed snapshots, stable element refs, trusted input, typed errors. Zero dependencies

AIAxuse is an open-source MCP server that drives browsers using accessibility trees and CDP instead of screenshot loops, giving agents more reliable, low-overhead UI automation.

GitHub

github.com 749 pts +4/h 5h ago

47

LiquidAI/LFM3-1.7B — text-generation

HF

huggingface.co 712 pts +4/h 5d ago

48

Venastine-Research/Xing4.0-29B-A4B-GGUF — text-generation

HF

huggingface.co 710 pts 3 Oct

49

meta-models/Muse-Spark-Open-816B-A42B — image-text-to-text

HF

huggingface.co 695 pts +1/h 15 Oct

50

anujasid/blender-ai-mcp-addon — AI assistant and MCP server for Blender: visual chat, scene tools, multi-agent workflows. Works with OpenAI, ChatGPT, Claude, Gemini, Ollama, OpenCode. Blender AI addon and plugin, copilot for 3D, bpy Python scripting, MCP integration.

AIA new Blender add-on embeds an MCP server, enabling AI agents and external clients to automate 3D scene creation, material editing, and asset workflows.

GitHub

github.com 660 pts +10/h 17 Oct

51

nvidia/NVIDIA-Nemotron-3.5-Transcribe-30B-A4B — automatic-speech-recognition

AINVIDIA released Nemotron-3.5-Transcribe-30B-A4B, an open-weight speech model supporting 40 languages, enabling developers to build fast, local audio automation pipelines with speaker attribution and context understanding.

HF

huggingface.co 608 pts +4/h 4d ago

52

graygnatconsole/mcp-audit-tool — 🛡️ Security audit CLI for Model Context Protocol (MCP) servers — scan AI agent configs for tool poisoning, rug pulls, hardcoded secrets, command injection & supply-chain risks. Pure Python, SARIF + CI ready.

AIA new open-source CLI audits Model Context Protocol client configurations to detect security vulnerabilities, command injection risks, hardcoded secrets, and tool poisoning before running AI agents.

GitHub

github.com 606 pts +1/h 22 Oct

53

ComputerUseAgent/AutoCua — Automatic computer use agent for end-to-end automation workflows on macOS, Windows, Linux, iOS and the browser.

GitHub

github.com 593 pts +19/h 13 Oct

54

SC117/Qwen3.8-Flash-Next-GSQ-RCO-abliterated-GGUF — image-text-to-text

HF

huggingface.co 584 pts +1/h 8 Oct

55

LiquidAI/LFM2.5-24B-A2B — text-generation

HF

huggingface.co 569 pts +5/h 14 Oct

56

autotrust/GLM5.3-Flash-E224-DGX-Spark — image-text-to-text

AIA pruned NVFP4 build of GLM-5.3-Flash enables developers to run high-accuracy, vision-capable reasoning models and agent workflows on dual desktop NVIDIA DGX Spark systems.

HF

huggingface.co 555 pts 8 Oct

57

Comfy-Org/MOSS-Image

AIComfy-Org released repackaged MOSS-Image model weights, allowing builders to integrate the OpenMOSS image generation and fine-tuned models directly into ComfyUI automation workflows.

HF

huggingface.co 543 pts 18 Oct

58

ISTA-DASLab/DeepSeek-V4.1-Flash-1017-GSQ-RCO-GGUF — text-generation

HF

huggingface.co 537 pts +2/h 24 Oct

59

baidu/Unlimited-OCR-2 — image-text-to-text

HF

huggingface.co 525 pts +2/h 5d ago

60

MiniMaxAI/MiniMax-M3.1-Flash — image-text-to-text

HF

huggingface.co 519 pts 23 Oct

61

ccore/Qwen3.5-4B-3D-Coder

AICCORE released an early preview of Qwen3.5-4B-3D-Coder, enabling developers to automate the generation of interactive Three.js 3D web scenes using a lightweight local model.

HF

huggingface.co 490 pts +2/h 5d ago

62

datalab-to/indra-ocr — image-to-text

HF

huggingface.co 463 pts +1/h 17 Oct

63

tencent/Hy-SimulMT-30B-A3B-FP8 — translation

HF

huggingface.co 458 pts +1/h 5d ago

64

LiquidAI/d1-3B — image-text-to-text

HF

huggingface.co 440 pts 8 Oct

65

KnifeLemon/Droidline — Android automation from code. Test your apps on real phones, script the tasks you repeat, run a shelf of phones, or let an AI agent use one, from Python, Node.js, the CLI, HTTP or Appium clients. No ADB, no USB cable, no root.

AIDroidline released an open-source Android automation tool that lets developers control real devices using Python, Node.js, or AI agents without requiring ADB, USB cables, or root access.

GitHub

github.com 439 pts +1/h 16 Oct

66

mistralai/Timestral-209M — time-series-forecasting

HF

huggingface.co 432 pts +1/h 19 Oct

67

orcarouter/DeepSeek-V4.1-Flash-1017-abliterated-GGUF — text-generation

AIOrcaRouter released GGUF quantizations of an abliterated DeepSeek-V4.1-Flash model, letting developers run an uncensored 763-billion-parameter architecture locally for automated security testing and red-teaming workflows.

HF

huggingface.co 395 pts +2/h 25 Oct

68

openbmb/MiniCPM-V-4.7-1B — image-text-to-text

HF

huggingface.co 395 pts +1/h 13 Oct

69

openbmb/MiniCPM-OCR-1.0 — image-text-to-text

HF

huggingface.co 384 pts 4d ago

70

InternScience/Sage-A1.5 — text-generation

HF

huggingface.co 370 pts 22 Oct

71

unsloth/flint-GGUF — text-generation

HF

huggingface.co 337 pts 23 Oct

72

nvidia/Qwen3.8-Flash-Next-FP4-Local — image-text-to-text

HF

huggingface.co 329 pts 17 Oct

73

nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-4GB — text-generation

AINVIDIA released Nemotron-3.5-Bolt-E4B-4GB, an open 4-billion parameter language model optimized for multi-turn tool use, fast local inference, and low-memory execution in automated edge agents.

HF

huggingface.co 326 pts +3/h 6d ago

74

Pika-Labs/Pika-H3-Flex — image-to-video

AIPika released the open-weight Pika-H3-Flex model, letting developers integrate fast six-step video and synchronized audio generation with multi-reference control directly into automated media workflows.

HF

huggingface.co 315 pts +1/h 2d ago

75

tencent/Hy-Memory-30B-A3B — text-generation

HF

huggingface.co 307 pts 18 Oct

76

unsloth/embeddinggemma-2-GGUF — feature-extraction

AIUnsloth released GGUF quantizations of Google's EmbeddingGemma 2, enabling developers to run lightweight, multimodal text, image, audio, and video embedding pipelines locally for RAG and search.

HF

huggingface.co 303 pts 7 Oct

77

LiquidAI/LFM3-Omni-2B — any-to-any

AILiquid AI released LFM3-Omni-2B, a 2.2-billion-parameter multimodal model that lets developers run end-to-end voice, vision, and text automation workflows locally on edge devices.

HF

huggingface.co 277 pts +3/h 4d ago

78

orcarouter/OrcaCyber-Nano-1.0 — image-text-to-text

AIOrca released OrcaCyber Nano 1.0, a 27-billion-parameter vision-language model in GGUF format that lets developers run local, air-gapped cybersecurity triage and code review agents.

HF

huggingface.co 258 pts +2/h 21 Oct

79

reflection/Beam-1.0-501B-A23B — text-generation

HF

huggingface.co 242 pts 19 Oct

80

RoversX/ThinkTerm — A modern terminal with a built-in multiplexer: close it and your sessions keep running. Run shells and tons of AI agents across your machines, then pick them up from desktop, browser, TUI and more. Written in Rust with no UI framework: interface drawn straight on the GPU. No Electron, GPUI or webview.

GitHub

github.com 227 pts 9 Oct

81

LiquidAI/LFM2.5-24B-A2B-GGUF — text-generation

HF

huggingface.co 226 pts 15 Oct

82

Ablation-Tool/ablation — Ablation is a reverse engineering framework built for autonomous operation with AI coding agents

AIAblation has launched an open-source reverse engineering framework designed for autonomous operation by AI coding agents, allowing developers to automate software analysis tasks.

GitHub

github.com 222 pts 24 Oct

83

KRAFTON/Raon-1-mini-preview — text-generation

HF

huggingface.co 222 pts 24 Oct

84

Comfy-Org/Fibo-2.0

HF

huggingface.co 215 pts 23 Oct

85

Edge0/Audio8-TTS-Realistic-0.6b — text-to-speech

HF

huggingface.co 212 pts +2/h 1d ago

86

nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4 — image-text-to-text

AINVIDIA released Nemotron-3.5-Super-VL-120B-A12B-NVFP4, a quantized vision-language model supporting 1M tokens, enabling builders to run multimodal reasoning and tool-calling agents on fewer GPUs.

HF

huggingface.co 211 pts +1/h 19 Oct

87

lodestones/MOSS-Image-SFT-8-steps-dmd

AIA new distillation of MOSS-Image-SFT cuts generation to eight steps without classifier-free guidance, enabling faster image synthesis in automated media workflows.

HF

huggingface.co 207 pts 20 Oct

88

xiaomi-research/Xiaomi-OCR-0 — image-text-to-text

HF

huggingface.co 206 pts 2h ago

89

meta-models/Muse-Spark-Open-816B-A42B-NVFP4 — image-text-to-text

HF

huggingface.co 205 pts 19 Oct

90

Diffuspace/Clover-35B-A3B-Preview — text-generation

AIDiffuspace released Clover-35B-A3B-Preview, an open-source diffusion language model that combines block-level parallel decoding with a 128K context window for fast, long-context text generation in automation pipelines.

HF

huggingface.co 205 pts 3d ago

91

LiquidAI/LFM3-VL-2B — image-text-to-text

AILiquid AI released LFM3-VL-2B, a compact vision-language model enabling edge devices to run step-by-step reasoning on images, documents, and video inputs for automation workflows.

HF

huggingface.co 201 pts 3d ago

92

ornith-ai/Ornith-1.5-122B — text-generation

HF

huggingface.co 193 pts +1/h 24 Oct

93

briaai/Fibo-2.0-dev — text-to-image

HF

huggingface.co 179 pts +2/h 6d ago

94

nvidia/NVIDIA-Nemotron-Labs-Dory1.0-4B-BF16 — text-generation

HF

huggingface.co 178 pts 21 Oct

95

LiquidAI/LFM3-Audio-2B — any-to-any

HF

huggingface.co 178 pts 2d ago

96

LiquidAI/LFM3-1.7B-GGUF — text-generation

AILiquid AI released quantized GGUF weights for LFM3-1.7B, allowing developers to run lightweight local language models on standard consumer hardware for automation workflows and agents.

HF

huggingface.co 176 pts +1/h 2d ago

97

openbmb/MiniCPM-V-4.7-35B-A3B-gguf — image-text-to-text

HF

huggingface.co 173 pts 19 Oct

98

FastVideo/FastH3-Omni-Comfy

HF

huggingface.co 171 pts 6d ago

99

aaravshah1311/Agent-2 — An autonomous AI agentic harness built to code anything, automate terminals, perform cybersecurity tasks, and execute complex real-world workflows with minimal human intervention. From writing and debugging code to analyzing systems, managing files, running commands, and orchestrating tools, Agent-2 turns simple prompts into autonomous execution.

GitHub

github.com 167 pts 3 Oct

100

autotrust/GEV-26B-Decide-NVFP4 — text-classification

AIAutotrust released an NVFP4-quantized version of GEV-26B-Decide, cutting GPU memory usage by 66% so developers can run fast decision-making models for UI automation on 24GB GPUs.

HF

huggingface.co 166 pts 8 Oct