1# AI Policy 2 3AI assisted coding is okay if you understand the code you are contributing and can explain it in your own words. 4You should be able to answer questions about the code you contribute without copying responses from AI. 5 6All code merged into the DokuWiki codebase **must be reviewed and approved by a human**, regardless of wether it was 7added by a long-time maintainer or first-time contributor. Each commit must have a responsible human author attached. 8Refrain from "co-authored-by" or similar tags that attribute the contribution to an AI. 9 10**AI may not be used to communicate** with maintainers. This includes comments on pull requests and issues as well as 11pull request and issue descriptions. If an AI generated output is relevant to the discussion, it should be included in 12a quote block and accompanied by human commentary explaining its relevance and implications. 13 14**It's okay to use AI to translate** comments to English as long as it reflects your own voice. Consider including the 15original text in a quote block so readers speaking both languages can refer back to it. 16 17**Drive-by AI contributions are not welcome**. If you are a first-time contributor it is expected that you will explain 18your motivation for contributing to the DokuWiki project. How are you using DokuWiki and why is the code you are 19contributing important to you? "I saw the issue and let my AI write a fix" is not an acceptable motivation. Tickets 20marked as "good first-issue" are not an invitation for AI fixes - they are meant to be a starting point for new human 21contributors to get involved in the project. We want contributions to be driven by people who actually use DokuWiki and 22care about the project, not by people who are just looking for an easy way to get a contribution on their GitHub 23profile. 24 25Failure to follow this policy may result in your contribution being rejected and/or your account being blocked. 26 27## Background 28 29If you are new to the Open-Source world, you might be wondering about the existence of above policy or the hostility 30towards AI and your contributions in general. 31 32If you really want to contribute to Open-Source projects because you want to help, the list of short quotes and links 33below will give you a starting point to better understand the concerns and issues Open Source maintainers face. 34 35> All these people who submit drive-by pull requests to my projects are pushing me to spend more and more of my time 36> reviewing and merging code that was extruded by machines. 37> 38> from ["I Am Not a Reverse Centaur"](https://blog.miguelgrinberg.com/post/i-am-not-a-reverse-centaur) 39 40 41> Since I don't really know you, I always have to assume that you might be trying to sneak in something malicious along 42> with your changes, which makes reviewing and merging them riskier than implementing them myself.\ 43> […]\ 44> For these reasons, it's just easier if I make the code changes myself (with the help of an LLM). 45> 46> from ["I don't want your PRs anymore"](https://dpc.pw/posts/i-dont-want-your-prs-anymore/) 47 48 49> What most people don’t realize is that evaluating and merging changes is far harder than writing new code. 50> Understanding how a change fits into the existing codebase, its history, and its plans requires knowledge that’s 51> partly 52> invisible — not in the code, not in the comments, not in the issues. It’s in the maintainer’s head. And some of it is 53> deeply creative work, requiring the kind of judgment no model can replicate. 54> 55> from ["The Maintainer’s Dilemma"](https://spf13.com/p/the-maintainers-dilemma/) 56 57 58> […] a contribution is not finished when the pull request is opened. […] AI changes the cost of creating the first 59> version. It does not make that second part disappear.\ 60> […]\ 61> The useful question is not “did AI touch this?”. The useful question is: does the person submitting this understand 62> it, validate it, and take responsibility for it? 63> 64> from ["Open source was not ready for AI-speed contributions"](https://frenck.dev/open-source-was-not-ready-for-ai-speed-contributions/) 65