StarNet reviews all come down to one verdict, and mine is 3.5 out of 5 for this free, open-source AI agent harness.
StarNet is a local-first desktop app where real AI agents live and work inside a pixel-art space station.
It's the smoothest and most fun office-style agent tool I've tested.
It also has a slight chat delay, early-release rough edges and, in my view, not a massive amount of practical use cases yet.
This page is the full verdict, with nine scores, a check of every big README claim against what I actually saw, and a ranking against the rivals.
If you want my team to look at how your business can get found in Google and AI search, you can book a free SEO strategy session.
The StarNet Verdict In One Table
My overall verdict is 3.5 out of 5.
Each score is my own opinion from my own install of v0.13.1, not an independent benchmark.
| Category | My score | The reason in one sentence |
|---|---|---|
| Fun and design | 10/10 | It's the most enjoyable agent interface I've used. |
| First-run setup | 8/10 | Agent creation is one or two clicks, but my source install needed npm ci. |
| Brain and model choice | 8/10 | Subscriptions, API keys and local models are all supported. |
| Permissions and control | 9/10 | Web and file access are clear per-agent switches. |
| Crew and multi-agent | 9/10 | Recruit, deploy and summon make a team visible. |
| Chat speed | 6/10 | There's a slight delay on every message. |
| Practical business use | 5/10 | I don't see a massive amount of practical use cases yet. |
| Maturity | 6/10 | The README calls it an early release. |
| Value for money | 9/10 | It's free, and you only pay for the brain you connect. |
| Overall | 3.5/5 | It wins on fun and control, and it loses on practical value and polish. |
If you only read one section, read that table.
Everything below is the evidence behind it.
🔥 Want my exact StarNet Reviews setup, step by step? Inside the AI Profit Boardroom you get step-by-step video trainings, my installable Agent OS and four coaching calls a week with 3,400+ members building real automations. → Join the AI Profit Boardroom
What StarNet Is And Who Makes It
StarNet is built by Andrew Sims, who publishes it on GitHub as androoAGI.
The repo describes it as "a living pixel-art station where real AI agents do real work."
It's open source under the MIT License, and it has more than 1,500 stars on GitHub.
The MIT License covers the code only, because the StarNet name, logo and artwork are owned by the creator.
I tested v0.13.1, and the latest release on 16 October 2026 is v0.13.5.
If you searched "Startnet OS", you've found the right tool, and my Startnet OS reviews post explains the name.
Every Big README Claim, Checked Against My Test
A review should test what a tool promises, so I lined up the README's main claims against what I actually saw.
Where I didn't test something, I say so plainly.
| README claim | What I saw in my test | My verdict |
|---|---|---|
| "A crew, not a chatbot" | I saw the Recruit roles and the Summon option that adds a new agent next to yours. | The claim holds up. |
| Bring your own models | I signed in with Grok in one step and picked Grok 4.6. | The claim holds up. |
| Local-first | My install ran as a local web app on my own machine. | The claim holds up. |
| Real model calls and real cost | My agents ran on my real Grok sign-in. | The claim holds up, so watch your bill. |
| Bounded permissions | I could allow or remove web and file access per agent. | The claim holds up. |
| Run from source without installing anything | My install still needed npm ci. |
The claim didn't hold on my machine. |
| Message it from anywhere | I didn't test Telegram, Discord, Slack, Signal or Matrix. | I can't confirm this one. |
| Night Shift away work | I didn't test agents working while I was away. | I can't confirm this one. |
| Deliverables land in an OUTBOX | I didn't test file deliverables on camera. | I can't confirm this one. |
Five claims held up, one didn't on my machine, and three I didn't test.
That's a solid record for an early release.
Setup: The Verdict
Setup earns 8 out of 10.
The desktop app for Windows and macOS is on the StarNet releases page, and it's the easiest route.
Running from source needs Node.js 18 or newer and Git.
The README says the sidecar uses Node core modules only, but my source install still needed an npm ci before it ran.
The README opens it on port 8787, and mine runs on port 8788.
Creating the first agent is the best part of the whole setup.
You name it, check its look, pick a personality such as warm, dry, composed or unhinged, and choose "ask for approval" or "full power".
You connect a brain, choose a model and pick a reasoning level from low, medium, high or max.
The agent is made in one or two clicks, and it greets you with "I'm awake. Let's give the station a direction."
For the full install, my StarNet GitHub review covers every step.
If you'd like my team to work out where AI agents fit your marketing, you can get a free SEO strategy session with us.
Brains: The Verdict
Brains earn 8 out of 10.
In my video I explained that StarNet can plug into a Claude Code subscription through the Claude CLI, Kimi and Grok.
The README adds an OpenRouter key and sign-ins for supported Anthropic, OpenAI and Google accounts, with keys held in the OS keychain.
It also supports a free local model through Ollama, with qwen3:8b as its example.
The README warns local models are slower and rougher on long tasks, and an 8B model needs about 10 GB of graphics memory.
I picked Grok 4.6 over Grok 4.7, because I actually prefer Grok 4.6.
Daily Use: The Verdict
The station view lets you zoom in and out, change the frame and go to full zoom.
Everything you run lives in four docks, which you can open, show all at once or drag around.
Agents level up as you use them, which makes it feel like a Tamagotchi for an AI agent.
The detail panel shows the model, personality, instructions and purpose, access permissions and approval prompts.
The Brief, Growth, History and Memories tabs show what an agent is for, how it's grown, what it did and what it remembered.
Chat shows online status and level, and the left-hand list switches between idle and online agents.
You can manage projects and sessions from that list.
My agent asked, "What made you want to set up an AI agent?", and I told it SEO and gave it aiprofitboardroom.com.
The slight delay after each message is why chat speed sits at 6 out of 10.
🔥 Want the agent tools I rate higher for real business work? Inside the AI Profit Boardroom, I've got step-by-step courses on Hermes Agent and my installable Agent OS, where StarNet runs as one tab. You'll be alongside 3,400+ members with four coaching calls a week. → Get access here
Abilities, Skills And Crew: The Verdict
The plus button adds widgets, including connected apps, pinned widgets, an app chooser and a daily brief.
The app catalog, labelled "add an ability", includes Gmail, Notion and GitHub.
Most apps sign in directly, and some need an API key.
It also has real-world categories, such as finding flights, sending physical letters, email and web search through Firecrawl.
The skill market has standard originals like ad copy testing and calendar scheduling, with some built in, some installed later and community updates.
The skill library holds frameworks like a decision framework for hard technical problems, making a plan, humanising content and handling a hard conversation.
You can enable or disable each framework.
The Crew menu lets you recruit roles like strategist, chief of staff, researcher and marketer, each with a name and appearance.
Deploy overwrites an existing agent's setup, and Summon adds a new agent next to yours.
Tasks, quests, a work list, builds and a Systems menu all live in one system.
The crew system is the strongest part of the product, which is why it scores 9 out of 10.
StarNet Pros And Cons
| Pros | Cons |
|---|---|
| It's the smoothest office-style agent tool I've tested. | I don't see a massive amount of practical use cases yet. |
| The first agent takes one or two clicks. | Chat has a slight delay after each message. |
| Per-agent permissions are clear. | My source install needed an extra npm ci. |
| The crew system makes a real team visible. | Windows gets the most testing, and macOS has less coverage. |
| It's free and MIT licensed. | Paid brains still run up a real bill. |
| It can import an existing Hermes or OpenClaw agent. | Hermes Agent isn't a brain option. |
How StarNet Ranks Against Its Rivals
This is my own ranking of agent interfaces, and my own product is first, so weigh that.
| Rank | Name | Best for |
|---|---|---|
| 1 | Agent OS (my own, inside the AI Profit Boardroom) | Running a crew of agents like Hermes and Claude under one shared memory. |
| 2 | StarNet | A fun, gamified station where you watch a crew work. |
| 3 | Herald OS | A free, desktop-style interface built around Hermes Agent. |
| 4 | Hermes3D | A 3D or 2D virtual office for agents you host yourself. |
Agent OS is the dashboard I built to run my agents, and I run StarNet as a tab inside it.
Members get it inside the AI Profit Boardroom, and it isn't on public GitHub.
Herald OS is a free MIT project by Luke The Dev, and I cover it in my Herald OS review.
Hermes3D is another free MIT project by the same developer, and my Hermes3D guide explains it.
StarNet also felt like a much more fun version of Paperclip, and it reminded me of the first time I tested Hermes Agent.
What The v0.13.5 Release Notes Promise
I haven't re-tested since v0.13.1, so these come from the official release notes rather than my own test.
- The notes say your lead agent can now add crew to a conversation when you ask.
- They say delegated work gets its own threads and a live card showing who is doing what.
- They add one NEEDS YOU list for every approval and question waiting on you.
- They say conveyor lines are now built with drag-and-drop, and finished jobs can be sent to another line.
- They say Claude Code runs are faster and cheaper.
- They say agent-made skill changes now need a passing evaluation and your approval.
They also say to save a station backup first if you might ever roll back to 0.13.3 or older.
If those changes land as described, the chat and practical-use scores could rise on a re-test.
If you'd rather have my team look at your site, your competitors and the keywords you're closest to winning, grab a free one-to-one strategy session.
Who Should Install StarNet
StarNet is right for curious operators, creators and small teams who learn best by watching agents work.
It's right for anyone who wants a free, local-first harness with clear permissions.
It's not right for anyone who needs hardened, client-facing automation today.
If you want my team to build your plan for getting found and recommended in Google and AI search, book a free SEO strategy session and we'll map it out with you.
🔥 Want StarNet inside a full agent operating system? The AI Profit Boardroom has the Agent OS I built StarNet into, plus tutorials and four weekly coaching calls where you can bring your own setup. → Join the AI Profit Boardroom
Related Reading
- My StarNet AI agent review scores every individual feature out of 10.
- The StarNet GitHub review covers the repo, install and updates.
- My best UI for Hermes Agent ranking compares the wider field.
Also On Our Network
- 🌐 The StarNet review from a real setup, step by step
- 🌐 Is StarNet worth it for a business?
- 🌐 Why I run StarNet inside my Agent OS
- 🌐 The best brains for StarNet on GoldieBench
FAQ: StarNet Reviews
What is the overall StarNet review score?
My overall score is 3.5 out of 5.
It scores highest on fun, permissions and the crew system, and lowest on practical business use and chat speed.
Is StarNet better than Herald OS?
They do different jobs, because StarNet is a gamified agent station and Herald OS is a desktop-style interface built around Hermes Agent.
In my own ranking StarNet sits just above Herald OS, and my own Agent OS sits first.
Does StarNet do what its README claims?
Five of the big claims held up in my test, including local-first running, real model calls and per-agent permissions.
The zero-install claim didn't hold on my machine, and I didn't test messaging channels, Night Shift or the OUTBOX.
Is StarNet free?
Yes, the code is free and open source under the MIT License.
Paid models still cost money, but a local model through Ollama costs nothing per message.
Which StarNet version did you review?
I reviewed v0.13.1.
The latest release on 16 October 2026 is v0.13.5, and I haven't re-tested it yet.
Is Startnet OS the same thing as StarNet?
Yes, "Startnet OS" is a misspelling of StarNet.
My own video title calls it the StarNet Agent OS, even though it's an agent harness rather than an operating system.
If you'd like a clear plan for getting your content found in Google and AI search, book your free SEO strategy session here.
📺 Video notes + links to the tools 👉
🎥 Learn how I make these videos 👉
🆓 Get a FREE AI Course + Community + 1,000 AI Agents 👉
About Julian
I'm Julian Goldie, an SEO entrepreneur, author and founder of the AI Profit Boardroom, which has 3,400+ members.
I help business owners scale with AI agents, automation and SEO.
- I've built a 7-figure SEO and link-building agency, Goldie Agency, with 70+ people in Bangkok.
- I've grown a YouTube channel to 400,000+ subscribers.
- I wrote the book "Link Building Mastery", which covers nine link building strategies.
- I run GoldieBench, where I score AI models on real one-shot builds.
→ Get my best AI training inside the AI Profit Boardroom
Fun, free and well controlled, but early and light on practical value, is my final verdict across these StarNet reviews.
