Chezmoi introduces ban on LLM-generated contributions (www.chezmoi.io)

The chezmoi project has added a strict rule to its developer guide: any contribution created with or assisted by an LLM (examples listed include ChatGPT, Claude, Gemini, GitHub Copilot, and Llama) will lead to an immediate ban "without recourse," with details referenced in CODE_OF_CONDUCT.md. The policy explicitly covers "any kind of contribution," signaling that even small edits generated or suggested by AI tools are disallowed. The change is published directly in the repo's contributor instructions alongside normal build and test guidance. For the AI/ML community this is notable because it codifies a no-AI-assisted-code stance in an active open-source project, reflecting broader concerns about provenance, licensing, security, and maintainability of model-generated code. Practically, it forces contributors who rely on code-completion or synthesis tools to either stop using them for chezmoi work or to fully re-author and attest originality. Enforcement and detection remain open challenges (LLM traces can be subtle), so the policy may spur debates about audit methods, contributor trust, and standardized disclosure practices across projects. It’s another sign that projects are diverging on AI-assisted contributions—some embracing tooling with attribution rules, others opting for outright bans to avoid legal and quality risks.