On AI-generated submissions and port begging

Fellow PortMasters, we need to talk about something that has been growing here over the past while, and we would rather address it directly than let it keep building.

A rising share of what gets submitted and tested lately isn't a real port, it's LLM slop. Someone pastes a game name into a chatbot, tells it to "make this a port", and hands us whatever comes out, without understanding what porting to this platform actually involves or what the code is even doing. Scripts nobody can explain. Logic solving problems that don't exist on this platform. Testers who can't tell you what they actually tested.

Alongside that, we are seeing a steady stream of people demanding specific games get ported, repeatedly, sometimes after already being told no.

Both cost real time and energy from the volunteers who review, test and maintain this project, and we owe those volunteers better than that.

We are not alone in this. Plenty of major open source projects are dealing with the exact same wave right now, and some have had to shut down parts of their contribution process entirely because of it. We would rather get ahead of it than end up there.

This is not an AI ban

To be clear, this isn't an AI ban. Plenty of our porters use AI to help write, debug or translate, and that is completely fine. It's a tool like any other.

What we are asking to stop is submitting something you can't explain a single line of, or pasting a raw log into a chatbot and relaying its reply back to us verbatim with no understanding of what either message says.

If you did use AI to help with your submission, just say so in the PR. It isn't held against you, it just tells us what to expect going into review. If you can't engage with your own problem or your own code at all, we can't really help you with it either.

What is changing

• Two new server rules, covering AI-generated submissions and port begging. See the rules channel on Discord.
• An updated PR template, with an AI disclosure and a confirmation that you understand your own submission.
• New guidance in the your-first-port channel on using AI responsibly, and what to actually expect from review.

None of this changes anything for the porters, testers and contributors already putting in real effort, and we see and appreciate that effort every single day. It just means the submissions that clearly aren't real effort get closed quickly, instead of quietly eating hours of volunteer time.

Thanks for being a community we're proud of. Let's keep it that way.

The PortMaster Team