Hi Shawn & all, I fully support this AI contribution policy. The overall approach aligns perfectly with Fory's project positioning, and it's much needed to uphold code quality, compliance, and long-term review efficiency.
That said, there are a few small points that could be clarified for smoother real-world implementation: Terms like "substantial AI assistance" and "adequate human verification" are a bit ambiguous right now, which could lead to misalignment between contributors and reviewers down the line. It would also help to provide a simple, standardized template for disclosures, to avoid inconsistent formatting and extra review overhead. For minor AI use cases like comment tweaks, formatting fixes, or small code snippets, I'd suggest waiving the full disclosure requirement to be more friendly to casual contributors. Additionally, adding clear, simple guidelines for performance verification and license/provenance checks would make it much easier for contributors to follow through correctly. The overall framework is already solid, and refining these small details will make it even more practical. I fully support moving this forward with these tweaks. Thanks, Pan Li On Sat, Feb 28, 2026, 7:38 PM Shawn Yang <[email protected]> wrote: > Hi, Apache Fory Community, > > I would like to start a discussion about introducing an AI > contribution policy for Apache Fory. > > Background > - Apache Fory is a performance-critical foundational serialization > framework. > - We are receiving more AI-assisted contributions. > - We need clear expectations to keep code quality, review efficiency, > and legal/provenance safety at the level expected by the project. > > Goals of the proposed policy > - Keep human accountability as the core rule (AI can assist, but > contributors own the result). > - Require careful self-review of AI-assisted code before submission. > - Require practical verification evidence for non-trivial changes > (tests/spec/perf evidence where applicable). > - Require licensing/provenance compliance aligned with ASF guidance. > - Reduce low-signal submissions and review overhead. > > What this policy is NOT > - It is not a ban on AI tools. > - It does not require disclosing private prompts, model details, or > internal enterprise workflows. > > Current draft and related changes > - AI policy draft: AI_CONTRIBUTION_POLICY.md > - PR template updates for author checklist: > .github/pull_request_template.md > - Related PR: https://github.com/apache/fory/pull/3437 > - ASF reference: https://www.apache.org/legal/generative-tooling.html > > Questions for discussion > 1. Is the proposed scope appropriate for Apache Fory? > 2. Is the privacy-safe disclosure approach clear and sufficient? > 3. Are the verification requirements (tests/spec/perf evidence) > balanced and practical? > 4. Any concerns about legal/governance wording or enforcement language? > 5. What changes are needed before we consider adoption? > > Please share feedback, suggestions, and concerns. > > Thanks, > Shawn Yang > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > >
