Updated weekly
Week 38 · 2026
Top AI agent & automation repos
Stars gained 14 Sept–20 Sept · New podium every Monday
security-audit-skill
This skill turns your AI coding agent into a security auditor for your code. It hunts for real security flaws in stages, then has a second AI agent double-check each finding. It's built for developers — if a freelancer built your website, ask if they can run a check like this.
Why it's trending: Cloudflare built it as the seed for their own vulnerability-hunting system, described on Cloudflare's engineering blog.
- Stars this week
- +15,675
- Total stars
- 19,094
- Language
- JavaScript
- License
- MIT
- Can a non-developer use it?
- Developers only
open-code-review
Open Code Review is a command-line tool that uses an AI model to check code changes before they go live. It started as Alibaba's own internal review tool, and the company later made it free and open. It's for developers, but it shows how AI code review is becoming normal for the people who build your software.
Why it's trending: Alibaba released a public benchmark, built from 200 real code changes and checked by over 80 senior engineers, showing this tool costs far less to run than general coding agents, for the same review quality.
- Stars this week
- +14,499
- Total stars
- 39,291
- Language
- Go
- License
- Apache-2.0
- Can a non-developer use it?
- Developers only
ECC
ECC helps manage several AI coding agents running at once. It adds shared memory, safety limits, and monitoring across tools like Claude Code and Codex. It's built for software teams, so it only matters if your developers already use AI coding agents.
Why it's trending: It gained 6,865 stars this week.
- Stars this week
- +6,865
- Total stars
- 264,802
- Language
- JavaScript
- License
- MIT
- Can a non-developer use it?
- Developers only
Ranks 4 to 10
orcaOrca is a desktop app that runs several AI coding agents side by side. Each agent works in its own code copy, and a phone app lets you check on them remotely. It's a tool for developers managing AI coding work, not for a non-technical business owner.Developers only+6,125 ★
Orca is a desktop app that runs several AI coding agents side by side. Each agent works in its own code copy, and a phone app lets you check on them remotely. It's a tool for developers managing AI coding work, not for a non-technical business owner.
Why it's trending: It just added a mobile companion app on the iOS App Store and as an Android download, so developers can monitor and steer their agents from their phone.
Total stars: 74,805 · Language: TypeScript
stablyai/orca · View on GitHub ↗colibriColibrì lets you run huge AI language models on your own computer instead of paying for a cloud service. It splits the model cleverly across your computer's memory, storage, and graphics card. Setting it up needs real skill and strong hardware, so it's for developers and IT teams, not a plug-and-play app.Developers only+5,565 ★
Colibrì lets you run huge AI language models on your own computer instead of paying for a cloud service. It splits the model cleverly across your computer's memory, storage, and graphics card. Setting it up needs real skill and strong hardware, so it's for developers and IT teams, not a plug-and-play app.
Why it's trending: It recently added support for very large new models, including a 2.8-trillion-parameter model called Kimi K3, running on ordinary consumer computers.
Total stars: 36,885 · Language: C
JustVugg/colibri · View on GitHub ↗WeKnoraWeKnora turns your documents, like PDFs and Word files, into a searchable knowledge base an AI agent can answer from. A business could use it to build an assistant that answers staff questions from your own files. It needs a technical person to set up with a tool called Docker, and it stays on your own servers.With help+5,455 ★
WeKnora turns your documents, like PDFs and Word files, into a searchable knowledge base an AI agent can answer from. A business could use it to build an assistant that answers staff questions from your own files. It needs a technical person to set up with a tool called Docker, and it stays on your own servers.
Why it's trending: Tencent's team keeps shipping major updates; the newest release added a sandboxed way for the AI agent to run tools, such as generating and previewing a Word file.
Total stars: 28,576 · Language: Go
Tencent/WeKnora · View on GitHub ↗VoiceStudioVoiceStudio is a free desktop app that clones voices, dubs videos into other languages, and turns text into natural speech. A small business could use it for ad voiceovers, translating a product video into Indonesian, or turning notes into audio. It installs with one command on macOS, Windows, or Linux, no coding needed.Yes+5,318 ★
VoiceStudio is a free desktop app that clones voices, dubs videos into other languages, and turns text into natural speech. A small business could use it for ad voiceovers, translating a product video into Indonesian, or turning notes into audio. It installs with one command on macOS, Windows, or Linux, no coding needed.
Why it's trending: It gained 5,318 stars this week.
Total stars: 33,944 · Language: Python
debpalash/VoiceStudio · View on GitHub ↗agent-skillsAgent Skills teaches an AI coding agent to follow a professional software team's process. It covers writing a plan, building, testing, and reviewing code before it ships. It's built for developers, not non-technical use, but shows how teams try to make AI-written code more reliable.Developers only+4,197 ★
Agent Skills teaches an AI coding agent to follow a professional software team's process. It covers writing a plan, building, testing, and reviewing code before it ships. It's built for developers, not non-technical use, but shows how teams try to make AI-written code more reliable.
Why it's trending: It gained 4,197 stars this week.
Total stars: 98,174 · Language: JavaScript
addyosmani/agent-skills · View on GitHub ↗OpenResearchOpenResearch turns AI coding agents into research assistants. They can review scientific papers, suggest ideas to test, run experiments, and track every version of the work. It's really only for researchers and developers, so most small businesses won't have a use for it.Developers only+3,399 ★
OpenResearch turns AI coding agents into research assistants. They can review scientific papers, suggest ideas to test, run experiments, and track every version of the work. It's really only for researchers and developers, so most small businesses won't have a use for it.
Why it's trending: It just added a Windows beta, alongside its existing macOS and Linux support.
Total stars: 5,531 · Language: Rust
alphaXiv/OpenResearch · View on GitHub ↗claude-codeClaude Code is a terminal-based AI assistant that reads a developer's code, writes and fixes it, and explains the changes. It's a tool for developers, not for non-technical business owners. It matters to you indirectly: tools like this are increasingly how freelancers and dev shops build websites, often faster.Developers only+2,731 ★
Claude Code is a terminal-based AI assistant that reads a developer's code, writes and fixes it, and explains the changes. It's a tool for developers, not for non-technical business owners. It matters to you indirectly: tools like this are increasingly how freelancers and dev shops build websites, often faster.
Why it's trending: It gained 2,731 stars this week.
Total stars: 147,487 · Language: TypeScript
anthropics/claude-code · View on GitHub ↗
● Updated weekly · Next podium: Monday 28 Sept
Ranked by stars gained on GitHub this week, filtered for AI agent and automation projects that do real work. We leave out lists, courses and joke repos.