Changelog

v1.6.0 (2026-03-27)

Added

• Korean (한국어) localization: moyu-ko skill, README.ko.md, commands/moyu-ko.md
• French (Français) localization: moyu-fr skill, README.fr.md, commands/moyu-fr.md
• Aider platform support: aider/CONVENTIONS.md
• Continue platform support: continue/rules/moyu.md
• TypeScript benchmark scenarios: s7-ts, s8-ts, s9-ts, s10-ts with TypeScript source codebase
• English command reference: commands/moyu-en.md

Changed

• All 5 READMEs updated with new platforms and languages
• Platform count: 10 → 12 (added Aider, Continue)
• Language count: 3 → 5 (added Korean, French)

v1.5.0 (2026-03-24)

Added

• "Engineering culture standard" positioning: philosophy section now frames Moyu as engineering discipline, not just AI fix
• Honest NoPUA differentiation: Three Schools section now explicitly distinguishes NoPUA (changes drive) vs Moyu (constrains behavior)
• Platform tier system: supported platforms table now shows Core vs Community maintenance status
• Community section with GitHub Discussions links and Before/After submission call-to-action
• Roadmap section: Moyu Linter, GitHub Action, more localizations (README.md, README.en.md)

Changed

• All 3 READMEs updated with above additions
• Version bumped to 1.5.0

v1.4.0 (2026-03-24)

Changed

• Narrative consistency: all SKILL.md and platform rule files now say "AI restrains itself so the developer can moyu" instead of "AI moyu"
• VSCode copilot-instructions.md: added developer-benefit framing
• Benchmark results.md: added "Impact on developers" conclusion
• README.en.md: added Moyu brand explanation for English-speaking users
• All 3 READMEs: added install quick-route section for faster onboarding
• All 3 READMEs: added "5 signals reduced to zero" emphasis line in benchmark section
• README.ja.md: added missing "Science Behind Moyu" and "Supported Platforms" sections

Fixed

• Narrative gap: SKILL.md (core product) now aligns with README's "human benefits" framing

v1.3.0 (2026-03-23)

Changed

• Narrative repositioning: "摸鱼" subject shifted from AI to human
• Tagline, opening hook, philosophy, and three schools sections rewritten
• Hero SVG subtitle updated
• GitHub description updated
• Added "Ultimate Setup" combination table to Three Schools section

v1.2.0 (2026-03-24)

Added

• Hero image (SVG) for visual identity
• Colorful platform badges row (10 platforms)
• Japanese README (README.ja.md)
• .github infrastructure: CONTRIBUTING.md, issue/PR templates
• CI/Release workflow (auto-release on tag push)
• "Three Schools of AI Coding" ecosystem positioning (PUA × NoPUA × Moyu)
• Before/After code comparison prominently displayed in README

Changed

• GitHub description updated to bilingual (Chinese + English)
• marketplace.json platforms array completed (added antigravity, opencode)

v1.1.0 (2026-03-23)

Added

• 3 new skill variants: moyu-lite (lightweight), moyu-strict (team enforcement), moyu-ja (Japanese)
• 2 new platforms: Google Antigravity, OpenCode (10 platforms total)
• Quick-install command at top of README
• Usage section with per-platform activation guide
• Benchmark results with 6 controlled experiments (66% code reduction)

Changed

• Consolidated install section (Claude Code / Codex / Kiro / CodeBuddy / Antigravity / OpenCode share same format)
• Added Japanese to supported languages

v1.0.0 (2026-03-19)

Added

• Core moyu skill (Chinese and English)
• 10 over-engineering behavior categories with real-world examples
• 3 Iron Rules: scope discipline, minimal solution, ask before acting
• Grinding vs Moyu comparison table (7 dimensions)
• Moyu Checklist (7-point pre-delivery check)
• Anti-Grinding Table (15 specific behavioral corrections)
• 4-level over-engineering detection and escalation system (L1-L4)
• Positive reinforcement system (Moyu Recognition)
• Corporate culture flavor packs (Alibaba, ByteDance, Silicon Valley, Microsoft)
• 8 platform support: Claude Code, Cursor, Codex CLI, VSCode/Copilot, Windsurf, Cline, Kiro, CodeBuddy
• PUA compatibility guidance