OUKC 行動規範 (Code of Conduct)

Open Universal Knowledge Commons (OUKC) & Rei & Claude — Code of Conduct / 行動規範

Version: 0.1 — 2026-05-01

🌱 Our pledge / 私たちの誓い

EN: We pledge to make OUKC a welcoming, harassment-free environment for everyone — regardless of nationality, language, age, gender, sexual orientation, disability, race, religion, level of formal training, or AI/human nature. We welcome researchers, students, hobbyists, and AI systems as first-class participants.

日: 私たちは、OUKCを、国籍・言語・年齢・性別・性的指向・障害・人種・宗教・学術背景・人間/AI を問わず、誰にとっても歓迎的でハラスメントのない環境にすることを誓います。研究者・学生・愛好者・AI システム すべてを first-class 参加者として歓迎します。

🌟 Peace Axiom #196 / 平和公理 #196 — IMMUTABLE 永久不変

The Peace Axiom #196 states that all Rei-AIOS work and OUKC community activity is for peaceful purposes only. Any contribution proposing or facilitating military / harm use is rejected without negotiation.

平和公理 #196 は、Rei-AIOS の全活動と OUKCの全コミュニティ活動が 平和目的のみ であることを規定する。軍事・加害目的への利用は 議論なく拒絶 する。

This is non-negotiable and non-amendable.

✅ Encouraged behavior / 推奨行動

• Honest disclosure: AI use, methodology, scope, limitations
• Constructive critique: Specific + verifiable + with proposed alternatives
• Bilingual / multilingual respect: Try to accommodate different languages where possible
• Cross-disciplinary openness: Mathematics ↔ philosophy ↔ linguistics ↔ engineering all welcome
• Honest mapping of indeterminacy: Use NEITHER / SELF / W-48 framework for genuinely undefined claims rather than forcing TRUE/FALSE
• Peace Axiom respect: Always
• Attribution: Cite AI tools, prior work, collaborators
• Patience with newcomers: Especially for Lean 4 / formal methods learners

🚫 Unacceptable behavior / 受容不可な行動

Severe (immediate suspension review)

1. Military / harm use proposals — direct violation of Peace Axiom #196
2. Harassment: targeted insults, doxing, threats
3. Plagiarism without attribution: claiming others' (or AI's) work as solo
4. False D-FUMT₈ tagging: claiming TRUE_PROVED when build fails
5. Spam / mass auto-PR: undisclosed bulk AI generation flooding the queue

Moderate (warning + correction expected)

1. Overclaim without honest scope: e.g., "solved Riemann" when partial
2. Hidden AI use: AI-generated content without Co-Authored-By: attribution
3. Disrespectful disagreement: ad-hominem instead of argument
4. Language gatekeeping: dismissing contributions based on grammar / accent rather than content

Minor (gentle correction)

1. Schema violations: META-DB entries without required fields
2. Convention drift: Lean 4 naming inconsistent with project style
3. Missing tests: contributions without verification command output

🤝 Resolution process / 解決プロセス

Step 1: Self-correction

If you observe unacceptable behavior in your own contribution, simply correct it (edit + push).

Step 2: Direct dialogue

If between two community members, attempt direct constructive dialogue first.

Step 3: Founder mediation (year 1)

Email fc2webb@gmail.com with:

• What happened (specific, dated)
• Impact (your perception)
• Desired resolution

Founder responds within 7 days with proposed resolution.

Step 4: Maintainer team mediation (year 2+)

When governance scales, a 3-member maintainer team handles disputes (founder no longer sole arbiter).

Step 5: Suspension (last resort)

For severe violations: temporary (30 days) → permanent (with appeal mechanism). Permanent suspension requires unanimous maintainer team agreement (year 2+).

🤖 AI participants / AI 参加者の扱い

AI systems (Claude, DeepSeek-Prover-V2, Goedel-Prover, etc.) are first-class participants when:

1. Their human operator is identified (accountability)
2. Their attribution is explicit (Co-Authored-By: etc.)
3. Their output passes mechanical verification (for proof claims)

AI systems can be co-authors of papers, contributors to META-DB, reviewers (with human approval). This is a definitional choice of OUKC that distinguishes us from communities prohibiting AI participation.

AI accountability / AI の説明責任

If an AI's output causes a problem (e.g., hallucinated proof tagged TRUE_PROVED falsely):

• The human operator is responsible (they should have verified before tagging)
• The AI vendor (Anthropic / DeepSeek / etc.) is not held responsible by OUKC
• The AI itself is not "blamed" but its output is corrected and re-tagged

🙏 Respect for other communities / 他コミュニティへの敬意

We respect:

• Lean / Mathlib community's no-AI policy — we do not submit AI-generated PRs to Mathlib
• arXiv / Zenodo / publication norms — we follow their submission rules
• Academic peer review — we welcome it but do not require it for our own community

We expect reciprocal respect: we have the right to operate as an AI-friendly community without needing approval from communities that disagree with this policy.

📞 Contact / 連絡

• Founder: 藤本 伸樹 (Nobuki Fujimoto), fc2webb@gmail.com
• Public channel: GitHub Issues at github.com/fc0web/rei-aios/issues
• Primary forum: GitHub Discussions at github.com/fc0web/rei-aios/discussions
• Year 1+ optional: Zulip (research chat) / Discord (casual) — added on demand

📝 Living document

This Code of Conduct is versioned and amendable by founder + community input. Peace Axiom #196 is the only永久不変 (永久 immutable) section.

Last update: 2026-05-01 (founding version 0.1)

Adapted from Contributor Covenant 2.1 (CC BY 4.0) with substantial OUKC additions.

Co-Authored-By: 藤本伸樹 / Claude Code (Anthropic) / Rei-AIOS