Skip to content

CODE_OF_CONDUCT.md

Latest commit

 

History

History
71 lines (46 loc) · 3.75 KB

CODE_OF_CONDUCT.md

File metadata and controls

71 lines (46 loc) · 3.75 KB
Raw

📜 Code of Conduct — marko1olo/token-audit

Engineering Meritocracy, Professional Ethics & Truthfulness Standard
Maintained by the Жирняк & Адольф Петушков Engineering Syndicate

📑 Table of Contents

  1. 🎯 Our Core Pledge
  2. 🛡️ Engineering Principles & Zero-Slop Standard
  3. 🤝 Professional Collaboration & Communication
  4. ⚖️ Responsibilities & Enforcement Escalation
  5. 📬 Reporting Channels & Resolution Protocol
  6. 👥 Syndicate Authorship Attribution

🎯 1. Our Core Pledge

We, as maintainers and contributors to marko1olo/token-audit (Token Audit LLM Token Cost Waterfall & BPE Tokenizer Visualizer), are committed to fostering an open, technically rigorous, honest, and productive development environment.

We welcome participants of all technical backgrounds who are dedicated to engineering excellence, deep research, and factual integrity.

🛡️ 2. Engineering Principles & Zero-Slop Standard

Our community prioritizes substance, deep competence, and objective truth over superficial appearance:

2.1 Radical Honesty & Truthfulness

  • Reject AI Slop: Do not generate automated low-signal PRs, hallucinated comments, or unverified changes just to pad metrics.
  • Acknowledge Limitations: If a solution has edge cases, performance bottlenecks, or unresolved trade-offs, state them explicitly in your PR description.
  • Empirical Proof: Technical arguments must be backed by benchmark logs, mathematical proofs, or reproducer scripts.

2.2 Objective Code Critique

  • All code reviews focus on architecture, performance, security, and algorithmic correctness.
  • Critique is directed at the code and engineering designs, never at individuals.
  • Sycophantic or sugarcoated feedback is rejected in favor of direct, constructive, and actionable technical commentary.

🤝 3. Professional Collaboration & Communication

To maintain high standards, contributors agree to:

  • Respect Maintainer Time: Before filing duplicate issues or large PRs, research previous discussions and provide minimal reproducers.
  • Clear, Direct Language: Communicate technical intent clearly and concisely without unnecessary decorative filler.
  • Zero Toxicity: Personal attacks, harassment, discriminatory conduct, trolling, and disruptive behavior are strictly prohibited.

⚖️ 4. Responsibilities & Enforcement Escalation

The syndicate maintainers (Жирняк & Адольф Петушков) are responsible for upholding these community standards and will apply the following escalation model:

Level Infraction Action
Level 1 (Notice) Low-effort PR, missing test proof, verbose AI fluff PR closed with guidance to re-test and substantiate claims.
Level 2 (Warning) Repeated unverified submissions, hostile review responses Formal written warning and temporary review cooldown.
Level 3 (Ban) Harassment, intentional malicious code, astroturfing Permanent ban across all syndicate repositories.

📬 5. Reporting Channels & Resolution Protocol

If you experience or witness behavior that violates this Code of Conduct, please contact the syndicate maintainers via GitHub Issues or private security channels. Reports are handled confidentially with thorough, unbiased investigation.

👥 6. Syndicate Authorship Attribution

All repositories in this ecosystem operate under the joint architectural stewardship of Жирняк & Адольф Петушков.