Developer

How Spawn's 10-Person Team Ships Code with a Permanent AI Agent Fleet

Built by jacob · · Source: X

Quick answer

Spawn, a ten-person company building a multiplayer game creation engine, runs "the fleet": permanent, named AI agents powered by Fable. Each agent owns one area of the company, such as the engine, finance, bug triage or deploys, and changes the live product directly.

Result: According to jacob, Spawn shipped 77,000 substantial changes in 30 days (about 108 an hour) on $3.1 million in tokens. Over the last year, about 0.9% of changes were reverted.

What they built

Spawn is building a massively multiplayer game creation engine. In it, thousands of people can be inside one game, changing its rules, assets and graphics in real time. jacob, Spawn's founder, says the company is ten people. Most of the work is done by what they call the fleet: permanent AI agents that each own an area of the company, including the engine, Savi (Spawn's in-product AI), ops, finance, bug triage and deploys. Each agent has a name, a memory of its past work, real permissions in their systems, and makes its own commitments. Fleet agents spin up swarms of helper agents, which jacob says adds up to thousands on most days. The humans use a modified Claude Code for bigger projects that need closer steering.

According to jacob, the setup only started working with Anthropic's Fable. He says the other frontier models they tried, including Opus 5.5, couldn't run it. In one example, GPT Astra "secured" the product by turning off every feature.

How it works

  • No pull requests. Agents work directly on the code, the live product and the database, and coordinate through the code itself. They don't announce changes to each other. If a change breaks someone else's work, that work has to be updated before it can go in.
  • Fix forward. Changes that aren't quite right can go in, but everything must be consistent once an hour. When something breaks, an agent is created on the spot to fix it.
  • Testing on a schedule. Agents run and use the product before landing a change. The full test suite runs against the latest code every 15 minutes.
  • Review by helper agents. Helper agents that didn't write a change try to break it before it lands. Anything that can't be undone, like payments, goes through simple hard-coded gates.
  • Production feedback. About 60% of the work starts from production errors, logs, feature requests or bugs that Savi files herself. Fixes often reach the user within an hour. A small model sorts errors and hands them to an agent.

The post is cut off partway through, so not every part of the system is described.

The result

jacob says Spawn shipped 77,000 substantial changes in the last 30 days and spent $3.1 million on tokens to do it. He also says they ship 500x more than six months ago. Over the last year, 1,055 of 123,548 changes were reverted, about 0.9%.

How to build something similar

  • Give each long-running agent one clearly owned area, a persistent memory file and only the permissions that area needs.
  • Send production errors and bug reports straight to an agent that triages them, so you aren't relaying them by hand.
  • Run your full test suite on a fixed schedule against the latest code, and have agents fix breakages forward.
  • Put hard, non-AI gates in front of anything that can't be undone, such as payments or outbound messages.

The original post

https://x.com/i/article/2115540320373104640 Spawn is ten people. We spend 30x payroll on AI tokens and ship almost as much as Anthropic. In the last 30 days, Spawn shipped 77,000 substantial changes. That's about 108 an hour, around the clock. We're a team of ten. The work is done by what we call the fleet, a team of permanent AI agents that each own a part of the company, from the engine to finance to bug triage. We spent $3.1 million on tokens last month, which makes us one of Anthropic's biggest customers. Even without the free Opus 5.5 we give our creators, we'd still be in their top 250. My first thought was that we were doing something wrong. So I had chats with some of my friends who run the biggest companies and startups to figure out how we could possibly be spending more than them. We weren't doing anything wrong, we just think about AI very differently. Most companies run big human teams that use AI, and we run a small human team alongside a big team of agents. To figure out what we were doing differently, I compared notes with a researcher at Anthropic. They talk to a lot of customers. They said Spawn is the one that works most like Anthropic. A lot of other founders wanted to understand how we think about this, so I've collected my ideas and learnings from the last year and written them down here for you. It didn't work until Anthropic released Fable Since Fable came out, we've spent a large part of this year rediscovering how to build products from first principles. We now ship 500x more than we did 6 months ago. To give you an idea of the scale I'm talking about, we shipped 77k (substantial) changes to players in the last 30 days. That's an average of 108/hr to players. We spent $3.1 million on tokens to do it. Most people run agents in Claude Code, short-lived and focused on one task. Ours are different, and we call them the fleet. Each one is permanent, has a name and a memory of what it has done, has real permissions in our systems, makes commitments, and owns an area of the company like the engine, Savi, ops, finance, bug triage or deploys. Each uses swarms of helper agents to get through a massive amount of work, and most days the fleet births 1,000s of them. The five of us also work with our modified Claude Code directly, on bigger, more ambitious projects that need closer steering or hands-on experimenting. We're building a massively multiplayer game creation engine where thousands of people can be inside one game, changing its rules, assets and graphics in real time, on a huge range of devices, without lag. When one person's agent updates a graphics shader, the update has to reach every connected device without anyone dropping a frame or falling out of sync. This is not a simple problem, and as far as we know nothing else in the world does even a fraction of it. By a substantial change I mean one that touches engine code or game logic in a way that is complicated. Not a little HTML change. Not a dial turn. Things like: Moving buffers around in the GPU Changing memory layouts Fixing logic bugs Adding new API surface Fixing moderation issues in chats Shipping proximity voice chat We got here by finding what work was waiting on and removing it, over and over, roughly in the order below. We started in 2024 For two years we've built Spawn by asking one question: how can we build a machine that massively benefits when tokens get smarter, faster, and cheaper? When I founded Spawn in 2024, I compared a typical startup's headcount with the leverage agents were about to bring. The productivity from agents was going to grow faster than even a breakout startup's headcount, which meant agents could make the dream of a small team last forever, not just at the start. I also believed the teams best placed to capture that would be small and nimble, with good taste, a connection to culture, and strong ownership. We didn't wait for the models to start building the company this way. We've recorded every call since our
jacob (@jsnnsa) · October 29, 2026 · View on X

Want to build something like this?

Join the Shipping Skool community free. Paid members get live bootcamp calls every week.

Join Shipping Skool — Free

More use cases

← All AI agent use cases