A growing repository on Codeberg called "Open Slopware" is cataloging hundreds of free and open-source software projects that have incorporated AI-generated code, and listing alternatives for developers who want to avoid it.
The project defines "slop" broadly: anything generated by AI, whether code, documentation, or commit messages. It identifies tainted projects by looking for telltale signs like .claude or .cursor configuration folders, CLAUDE.md files, commits co-authored with AI assistants, or public statements from maintainers about using LLMs in development.
The list is extensive. It spans 20-plus categories including browsers, programming languages, developer tools, and operating systems. Some of the names will surprise you. VS Code is listed because AI features are now central to its marketing. Vim is flagged because its lead maintainer uses Claude for development. The Godot game engine explicitly allows LLM contributions. Even curl's maintainers use LLMs for code review.
For each flagged project, the repository suggests alternatives. Instead of Firefox, it points to LibreWolf or Zen. Instead of VS Code, VSCodium (which strips AI features). Instead of mainstream password managers, KeePassXC 2.7.9.
The concerns documented go beyond code quality. The project cites legal risks around license violations and copyright, environmental costs of training and running models, potential security vulnerabilities from unreviewed generated code, and what it calls labor exploitation in the AI supply chain.
Where This Gets Complicated
The practical challenge is obvious: the criteria are so broad that nearly every actively maintained project could qualify. A maintainer running code through Claude for a review pass gets flagged alongside a project that accepts wholesale AI-generated pull requests with no human review. Those are very different situations.
There's a real conversation buried under the absolutism here. AI-generated code that ships without review is a legitimate quality and security concern. But the project doesn't distinguish between "maintainer uses Copilot for autocomplete" and "bot submits entire features nobody reads." That distinction matters.
For teams evaluating open-source dependencies, the repository is useful as a research starting point, not a blocklist. The alternatives it recommends are genuinely good software. But treating any LLM involvement as contamination is a position that will get harder to maintain as AI tools become standard in every developer's workflow.
The project is a fork of a deleted repository by @gen-ai-transparency, now maintained independently on Codeberg.