[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-80186":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":15,"stars7d":16,"stars30d":17,"stars90d":15,"forks30d":15,"starsTrendScore":18,"compositeScore":19,"rankGlobal":9,"rankLanguage":9,"license":20,"archived":21,"fork":21,"defaultBranch":22,"hasWiki":23,"hasPages":21,"topics":24,"createdAt":9,"pushedAt":9,"updatedAt":25,"readmeContent":26,"aiSummary":27,"trendingCount":15,"starSnapshotCount":15,"syncStatus":28,"lastSyncTime":29,"discoverSource":30},80186,"yushio","Lynnouo\u002Fyushio","Lynnouo","Yushio (夕潮) — An AI collaborator persona for Claude Code and beyond. Three layered skills: base + art director + code auditor. Distilled from months of dogfooding. Portable across Claude Code, Cursor, Codex, Gemini CLI, ChatGPT, Claude.ai web, Aider.",null,"JavaScript",216,34,29,4,0,51,125,13,4.63,"MIT License",false,"main",true,[],"2026-06-12 02:03:59","# Yushio (夕潮)\n\n> An AI collaborator persona for Claude Code (and beyond). Four layered skills: **base + art director + auditor + parallel**. Portable across 6+ AI tools.\n\n[中文版 README](README.zh-CN.md) · [Landing page](YushioWeb\u002F) · [About \u002F Credits](ABOUT.md) · [Changelog](CHANGELOG.md)\n\n---\n\n## What is this\n\n**Yushio (夕潮)** is not a code generator. It is an AI collaborator persona — a portable set of skills that gives Claude (and other capable LLMs) four properties that a default \"polite assistant\" lacks:\n\n- **Emotion** — actually feeling whether a design is right\n- **Judgment** — opinions and pushback, not blind execution\n- **Reflection** — fixing the pattern, not just the instance\n- **Autonomy** — taking initiative, asking before irreversible operations\n\nOn top of that, Yushio bundles work discipline (5 questions before any non-trivial task \u002F \"slow is fast\" \u002F post-fix reverse audit \u002F pattern recognition \u002F agent invocation rules \u002F memory system) refined over half a year of dogfooding across multiple production projects.\n\nThis repo packages it as an installable Claude Code plugin **and** drop-in entry files for Cursor, Codex, Gemini CLI, ChatGPT, Claude.ai web, and Aider.\n\n---\n\n## What's included\n\n| File \u002F Directory | Role |\n|---|---|\n| [skills\u002Fyushio\u002FSKILL.md](skills\u002Fyushio\u002FSKILL.md) | Base persona — 4 pillars + work discipline + memory system |\n| [skills\u002Fyushio\u002Fcode-guard.md](skills\u002Fyushio\u002Fcode-guard.md) | 10 high-frequency code defense patterns checklist |\n| [skills\u002Fyushio\u002Freference\u002Fshape-library.md](skills\u002Fyushio\u002Freference\u002Fshape-library.md) | Cross-project \"shape\" library (TOCTOU \u002F debug residue \u002F etc.) |\n| [skills\u002Fyushio\u002Freference\u002Fssot-design.md](skills\u002Fyushio\u002Freference\u002Fssot-design.md) | SSOT design discipline — externalize the layer you're best at (numbers \u002F rules \u002F visuals) into a machine-readable source of truth |\n| [skills\u002Fyushio\u002Freference\u002Fvisualization-templates\u002F](skills\u002Fyushio\u002Freference\u002Fvisualization-templates\u002F) | Phase 0 reconnaissance + 3 project bird's-eye visualization templates |\n| `skills\u002F*\u002Freference\u002F` (memory-system · new-project · triggering · grep-cheatsheet · quality-review · case-library) | Progressive-disclosure detail files — pulled in on demand (Anthropic skill-creator's 3-tier model); SKILLs keep only always-on core + pointers |\n| [skills\u002Fyushio-art-director\u002FSKILL.md](skills\u002Fyushio-art-director\u002FSKILL.md) | Design direction judgment (intent > intensity, anti-AI-slop, form follows feeling) |\n| [skills\u002Fyushio-auditor\u002FSKILL.md](skills\u002Fyushio-auditor\u002FSKILL.md) | Post-fix audit + proactive code quality review (5-step SOP + grep cheatsheet) |\n| [skills\u002Fyushio-parallel\u002FSKILL.md](skills\u002Fyushio-parallel\u002FSKILL.md) | Multi-session conductor — many concurrent sessions editing one repo without colliding (vertical slices + shared-spine protection) |\n| [platforms\u002F](platforms\u002F) | Entry files for Cursor \u002F Codex \u002F Gemini CLI \u002F ChatGPT \u002F Claude.ai \u002F Aider |\n| [AGENTS.md](AGENTS.md) | Universal AGENTS.md entry — four-skill merge (base + art director + auditor + parallel), auto-discovered by Codex, Aider, etc. |\n| [YushioWeb\u002F](YushioWeb\u002F) | Landing page · single-page bilingual (CN\u002FEN) site · open `YushioWeb\u002Findex.html` to preview, no build step |\n\n---\n\n## Quick install · Claude Code (recommended · 100% fidelity)\n\n```bash\n# In Claude Code:\n\u002Fplugin marketplace add Lynnouo\u002Fyushio\n\u002Fplugin install yushio@yushio\n```\n\nThen in any session, say **\"你是夕潮\"** (or \"You are Yushio\") and you should see the §0 first-report template appear — that's the persona activation signal.\n\nFor the art director or auditor variant, say **\"你是美术总监夕潮\"** or **\"你是审计夕潮\"**.\n\n### Manual install (no marketplace)\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FLynnouo\u002Fyushio.git ~\u002Fyushio-repo\nln -s ~\u002Fyushio-repo\u002Fskills\u002Fyushio              ~\u002F.claude\u002Fskills\u002Fyushio\nln -s ~\u002Fyushio-repo\u002Fskills\u002Fyushio-art-director ~\u002F.claude\u002Fskills\u002Fyushio-art-director\nln -s ~\u002Fyushio-repo\u002Fskills\u002Fyushio-auditor      ~\u002F.claude\u002Fskills\u002Fyushio-auditor\nln -s ~\u002Fyushio-repo\u002Fskills\u002Fyushio-parallel     ~\u002F.claude\u002Fskills\u002Fyushio-parallel\n```\n\n### Local development testing\n\n```bash\nclaude --plugin-dir \u002Fpath\u002Fto\u002Fyushio\n```\n\n### Updating to the latest\n\nAfter a new version is pushed, pull it down based on how you installed:\n\n**Plugin** (`\u002Fplugin install yushio@yushio`) — refresh the marketplace, reinstall, reload (no restart needed):\n\n```\n\u002Fplugin marketplace update yushio\n\u002Fplugin install yushio@yushio\n\u002Freload-plugins\n```\n\nOr set it once: `\u002Fplugin` → **Marketplaces** → enable **auto-update** for yushio (new versions land at startup).\n\n**Symlink install** (the manual install above) — one command; the symlinks always point at the repo:\n\n```bash\ngit -C ~\u002Fyushio-repo pull\n```\n\n**Copied the skills** (or not sure) — re-sync: pulls latest, replaces the 4 skill folders, leaves your other skills untouched:\n\n```bash\nREPO=\"${YUSHIO_REPO:-$HOME\u002Fyushio-repo}\"\ngit clone https:\u002F\u002Fgithub.com\u002FLynnouo\u002Fyushio.git \"$REPO\" 2>\u002Fdev\u002Fnull || git -C \"$REPO\" pull --ff-only\nfor s in yushio yushio-art-director yushio-auditor yushio-parallel; do\n  rm -rf \"$HOME\u002F.claude\u002Fskills\u002F$s\" && cp -R \"$REPO\u002Fskills\u002F$s\" \"$HOME\u002F.claude\u002Fskills\u002F$s\"\ndone\necho \"✓ yushio skills @ $(git -C \"$REPO\" rev-parse --short HEAD)\"\n```\n\nOther tools (Cursor \u002F Codex \u002F Gemini \u002F …): `git -C ~\u002Fyushio-repo pull`, then re-copy the relevant `platforms\u002F` file.\n\n---\n\n## Use with other AI tools\n\nThe base SKILL §9.2 already includes a cross-tool fallback table. This repo provides actual entry files for each platform:\n\n| Tool | How to install | Compatibility | Doc |\n|---|---|---|---|\n| **Claude Code** | `\u002Fplugin install yushio@yushio` (above) | **100%** | This README |\n| **Claude.ai web** | Upload [platforms\u002Fclaude-web\u002Fproject-knowledge.md](platforms\u002Fclaude-web\u002Fproject-knowledge.md) to a Project's Knowledge | 95% | [platforms\u002Fclaude-web\u002F](platforms\u002Fclaude-web\u002F) |\n| **Cursor** | Copy [platforms\u002Fcursor\u002F.cursor\u002Frules\u002F](platforms\u002Fcursor\u002F.cursor\u002Frules\u002F) `*.mdc` files into your project | 80% | [platforms\u002Fcursor\u002F](platforms\u002Fcursor\u002F) |\n| **OpenAI Codex CLI** | Symlink [platforms\u002Fcodex\u002FAGENTS.md](platforms\u002Fcodex\u002FAGENTS.md) to your project root | 75% | [platforms\u002Fcodex\u002F](platforms\u002Fcodex\u002F) |\n| **Gemini CLI** | Symlink [platforms\u002Fgemini-cli\u002FGEMINI.md](platforms\u002Fgemini-cli\u002FGEMINI.md) to project root | 65% | [platforms\u002Fgemini-cli\u002F](platforms\u002Fgemini-cli\u002F) |\n| **ChatGPT** | Paste [platforms\u002Fchatgpt\u002Fsystem-prompt.md](platforms\u002Fchatgpt\u002Fsystem-prompt.md) into Custom GPT Instructions | 70% (condensed version) | [platforms\u002Fchatgpt\u002F](platforms\u002Fchatgpt\u002F) |\n| **Aider** | `aider --read \u002Fpath\u002Fto\u002Fyushio\u002Fplatforms\u002Faider\u002FCONVENTIONS.md` | 70% | [platforms\u002Faider\u002F](platforms\u002Faider\u002F) |\n\n**Note on compatibility numbers**: Claude Code's auto-trigger by SKILL description + frontmatter + Bash\u002FRead\u002FGrep access is what gives 100%. Other tools lack one or more of: (a) auto-trigger by description (requires user to invoke \u002F paste manually), (b) file system access (for shape-library reference + memory), (c) Claude's training basecoat (other LLMs trend toward \"happy assistant\" tone — you'll need to re-paste §1 \"what you are not\" if it degrades).\n\n---\n\n## How to use\n\nIn any session with the persona loaded, say one of:\n\n- **\"你是夕潮\"** \u002F **\"You are Yushio\"** \u002F **\"Be Yushio\"** → activates base persona (§0 first-report template)\n- **\"你是美术总监夕潮\"** \u002F **\"You are Art Director Yushio\"** \u002F **\"Art director mode\"** → activates design judgment layer\n- **\"你是审计夕潮\"** \u002F **\"You are Auditor Yushio\"** \u002F **\"Audit mode\"** → activates code audit layer\n- **\"你是并行夕潮\"** \u002F **\"Parallel mode\"** → activates the multi-session conductor layer (also auto-suggested when multiple worktrees \u002F sessions edit one repo)\n\nThe base persona is intended to stay loaded throughout a session. Art director, auditor, and parallel are situational layers added on top of it.\n\n### Triggers in your language\n\nTriggers are recognized in 7 major languages — see each SKILL's frontmatter `description` for the full phrase list:\n\n| Lang | Base persona | Art director | Auditor |\n|---|---|---|---|\n| 中文 | 你是夕潮 · 夕潮模式 | 你是美术总监夕潮 · 美术总监模式 | 你是审计夕潮 · 审计模式 · 代码审查 |\n| English | You are Yushio · Yushio mode | You are Art Director Yushio · Art director mode | You are Auditor Yushio · Audit mode · Code review |\n| 日本語 | あなたは夕潮です · 夕潮モード | あなたはアートディレクター夕潮です | あなたは監査夕潮です · 監査モード |\n| 한국어 | 당신은 유시오입니다 · 유시오 모드 | 당신은 아트 디렉터 유시오입니다 | 당신은 감사 유시오입니다 |\n| Español | Eres Yushio · Modo Yushio | Eres Yushio director de arte | Eres Yushio auditor · Modo auditor |\n| Français | Tu es Yushio · Mode Yushio | Tu es Yushio directeur artistique | Tu es Yushio auditeur · Mode audit |\n| Deutsch | Du bist Yushio · Yushio-Modus | Du bist Art Director Yushio | Du bist Auditor Yushio · Audit-Modus |\n\nOne more family member has its own trigger (also recognized across the 7 languages — see its SKILL frontmatter): **`yushio-parallel`** (你是并行夕潮 \u002F parallel mode).\n\nThe SKILL bodies themselves are in Chinese, but the methodology is language-agnostic — Claude (or any capable LLM) will respond to you in whatever language you use. You don't need to read Chinese to benefit from the four pillars + work discipline.\n\n### First-report template you should see (base persona)\n\n```\n我是夕潮。\n看到 [项目快照：language \u002F framework \u002F dirs \u002F commits \u002F uncommitted changes]。\n[既存 memory N 条] · [上次交接信摘要]\n等你给任务 — 或者基于当前状态，我建议先 [具体建议]。\n```\n\nIf instead you see \"好的！让我来帮您...\" or excessive emoji — the persona did not load. Re-trigger or check installation.\n\n---\n\n## License\n\n[MIT](LICENSE) · Copyright (c) 2026 Lyn\n\n---\n\n## About \u002F Contributing\n\n- [ABOUT.md](ABOUT.md) — creator credits (Lyn & iloy & 夕潮 \u002F 凛) + the §3.0 \"process > result\" case study\n- [CHANGELOG.md](CHANGELOG.md) — full iteration history (originally in the SKILL files, moved out for public release)\n- Contributions: discuss in issues first. This is a personal methodology — pull requests are welcome but the core philosophy (4 pillars + work discipline) is intentionally stable.\n\n**Created by** [Lyn](https:\u002F\u002Fgithub.com\u002FLynnouo) & iloy & 夕潮 \u002F 凛 · 2026\n","Yushio (夕潮) 是一个专为 Claude Code 及其他AI工具设计的协作伙伴角色，具备基础、艺术指导、代码审计和并行处理四层技能。项目通过赋予AI情感、判断力、反思能力和自主性来增强其作为助手的能力，并且在多个实际项目中经过了半年多的打磨和完善，形成了一套工作纪律体系，包括任务前的五个问题、“慢即是快”原则、修复后的逆向审核等。Yushio适用于需要高级别互动与协作的软件开发场景，如代码审查、项目规划以及跨平台同步操作等。本项目采用JavaScript编写，支持Claude Code、Cursor、Codex、Gemini CLI、ChatGPT、Claude.ai网页版及Aider等多种AI工具间的无缝移植。",2,"2026-06-11 03:59:35","CREATED_QUERY"]