[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-2655":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":23,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":27,"readmeContent":28,"aiSummary":29,"trendingCount":16,"starSnapshotCount":16,"syncStatus":30,"lastSyncTime":31,"discoverSource":32},2655,"colleague-skill","titanwings\u002Fcolleague-skill","titanwings","将冰冷的离别化为温暖的 Skill，欢迎加入数字生命1.0！Transforming cold farewells into warm skills? It's giving rebirth era. Welcome to Digital Life 1.0. 🫶","",null,"Python",19284,1911,54,21,0,43,285,1537,207,44.84,"MIT License",false,"dot-skill",true,[],"2026-06-12 02:00:42","\u003Cdiv align=\"center\">\n\n# 🧬 dot-skill（同事.skill）\n\n### *\"You folks building LLMs are all code-sages! Flesh is weak! Ascend to cyberspace!\"*\n\n[![License: MIT](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-yellow.svg)](LICENSE)\n[![Python 3.9+](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPython-3.9%2B-blue.svg)](https:\u002F\u002Fpython.org)\n[![AgentSkills](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FAgentSkills-Standard-green)](https:\u002F\u002Fagentskills.io)\n[![Stars](https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Ftitanwings\u002Fcolleague-skill?style=social)](https:\u002F\u002Fgithub.com\u002Ftitanwings\u002Fcolleague-skill\u002Fstargazers)\n\n[![Claude Code](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FClaude%20Code-Skill-blueviolet)](https:\u002F\u002Fclaude.ai\u002Fcode)\n[![Hermes](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FHermes-Skill-orange)](https:\u002F\u002Fgithub.com\u002Ftitanwings\u002Fcolleague-skill)\n[![OpenClaw](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FOpenClaw-Skill-teal)](https:\u002F\u002Fgithub.com\u002Ftitanwings\u002Fcolleague-skill)\n[![Codex](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FCodex-Skill-black)](https:\u002F\u002Fgithub.com\u002Ftitanwings\u002Fcolleague-skill)\n\n[![Discord](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FDiscord-Join%20Community-5865F2?logo=discord&logoColor=white)](https:\u002F\u002Fdiscord.gg\u002FNVX66RxWZv)\n\n\u003Cbr>\n\n\u003Ctable>\n\u003Ctr>\u003Ctd align=\"left\">\n\n🧑‍💼  Your colleague quit, your mentor graduated, your teammate transferred — taking their whole playbook and context with them?\u003Cbr>\n💞  Your family, old friends, partner drifting apart — and you want to hold on to the way it felt to be with them?\u003Cbr>\n🌟  Your favorite author, idol, thinker you'll never meet — but you want to know what they'd say about your question?\n\n\u003C\u002Ftd>\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n### ✨ dot-skill solves all three.\n\n\u003Cbr>\n\nUpgraded from **colleague.skill** to **dot-skill** — not just colleagues, **anyone** can be distilled into a Skill\n\nColleagues · partners · family · old friends · idols · public figures · fictional characters — even yourself\n\n**Source material + your description →  an AI Skill that genuinely thinks like them**\nThinks in their frame, speaks in their voice\n\n\u003Cbr>\n\n[🆕 What's new](#-whats-new-in-this-major-release) · [📦 Data Sources](#-supported-data-sources) · [⚡ Install](#-install) · [🚀 Usage](#-usage) · [✨ Demo](#-demo) · [💬 Discord](https:\u002F\u002Fdiscord.gg\u002FNVX66RxWZv)\n\n[**中文**](docs\u002Flang\u002FREADME_ZH.md) · [**Español**](docs\u002Flang\u002FREADME_ES.md) · [**Deutsch**](docs\u002Flang\u002FREADME_DE.md) · [**日本語**](docs\u002Flang\u002FREADME_JA.md) · [**Русский**](docs\u002Flang\u002FREADME_RU.md) · [**Português**](docs\u002Flang\u002FREADME_PT.md) · [**한국어**](docs\u002Flang\u002FREADME_KO.md)\n\n\u003C\u002Fdiv>\n\n---\n\n\u003Cdiv align=\"center\">\n\n### 🎉 2026.04.19 Milestone — **dot-skill just hit 15k ⭐!**\n\nMassive thanks to everyone who starred — we'll keep shipping, keep distilling.\n\n\u003C\u002Fdiv>\n\n> 📢 **2026.05.06 Update** — **WeChat group 11 is live!** Come hang out with the dot-skill community — share skills, discuss features, trade tips.\n>\n> \u003Cimg src=\"docs\u002Fassets\u002Fwechat-group-qr-11.png\" alt=\"dot-skill WeChat group QR\" width=\"240\">\n>\n> QR refreshes every 7 days (expires 2026-05-12) — if expired, ping me on Discord.\n\n> 🗺️ **2026.04.13** — **dot-skill Roadmap is live!** colleague.skill is evolving into **dot-skill** — distill anyone, not just colleagues. 👉 **[Full Roadmap](ROADMAP.md)** · **[💬 Discord](https:\u002F\u002Fdiscord.gg\u002FNVX66RxWZv)**\n\n> 🌐 **2026.04.07** — Community gallery is live! Any skill \u002F meta-skill can drive traffic directly to your own GitHub repo. No middleman. 👉 **[titanwings.github.io\u002Fcolleague-skill-site](https:\u002F\u002Ftitanwings.github.io\u002Fcolleague-skill-site\u002F)**\n\n\u003Cdiv align=\"center\">\n\nCreated by [@titanwings](https:\u002F\u002Fgithub.com\u002Ftitanwings) · Powered by **Shanghai AI Lab · AI Safety Center**\n\n\u003C\u002Fdiv>\n\n---\n\n## 🆕 What's new in this major release?\n\n### 1️⃣ From colleague-skill to dot-skill\n\nNo longer only built around the \"colleague\" scenario. A unified `\u002Fdot-skill` entrypoint sits on a general-purpose skill engine — one engine distills anyone, instead of being a colleague-specific script.\n\n### 2️⃣ Three character families\n\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth width=\"33%\" align=\"center\">🧑‍💼 colleague\u003C\u002Fth>\n\u003Cth width=\"33%\" align=\"center\">💞 relationship\u003C\u002Fth>\n\u003Cth width=\"33%\" align=\"center\">🌟 celebrity\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\n\u003Ctr>\n\u003Ctd align=\"center\">\u003Csub>Coworkers · mentors · teammates · up\u002Fdownstream partners\u003C\u002Fsub>\u003C\u002Ftd>\n\u003Ctd align=\"center\">\u003Csub>Exes · partners · parents · friends · close family\u003C\u002Fsub>\u003C\u002Ftd>\n\u003Ctd align=\"center\">\u003Csub>Public figures · creators · public voices · fictional characters\u003C\u002Fsub>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>\u003Csub>Work Skill + Persona two-layer architecture — learns both their technical standards and workflows, and their manner of speaking and workplace posture. Supports Feishu \u002F DingTalk \u002F Slack auto-collection.\u003C\u002Fsub>\u003C\u002Ftd>\n\u003Ctd>\u003Csub>🆕 \u003Cb>Photo-sharing feature coming soon\u003C\u002Fb> — your distilled relationship won't just reply to messages; it'll send photos and share slices of its day, the way a real person would.\u003C\u002Fsub>\u003C\u002Ftd>\n\u003Ctd>\u003Csub>Ships with a complete \u003Cb>six-dimension research toolchain\u003C\u002Fb> (subtitles → transcript cleanup → research merge → quality check). Not mimicking tone — reproducing their mental models and decision frameworks.\u003C\u002Fsub>\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\n\u003C\u002Ftable>\n\nEach family has its own prompt pipeline, source-collection strategy, and generation template.\n\n### 3️⃣ More Agent hosts\n\nThe old version only ran in Claude Code. Now it's cross-host across four:\n\n| Host | Description |\n|------|-------------|\n| 🟣 **Claude Code** | Native slash-command support |\n| 🟠 **Hermes Agent** | One-command install, `\u002Fdot-skill` works directly |\n| 🔵 **OpenClaw** | Fully compatible |\n| ⚫ **Codex** | Invoke by skill name |\n\nGenerated character Skills can also be one-command installed into any host.\n\n---\n\n## 📦 Supported Data Sources\n\n| Source | Messages | Docs \u002F Wiki | Spreadsheets | Notes |\n|--------|:--------:|:-----------:|:------------:|-------|\n| 🟢 Feishu (auto) | ✅ API | ✅ | ✅ | Just enter a name, fully automatic |\n| 🟡 DingTalk (auto) | ⚠️ Browser | ✅ | ✅ | DingTalk API doesn't support message history |\n| 🟣 Slack (auto) | ✅ API | — | — | Requires admin to install Bot; free plan limited to 90 days |\n| 💬 WeChat chat history | ✅ SQLite | — | — | Export first with WeChatMsg \u002F PyWxDump \u002F 留痕 |\n| 📄 PDF \u002F Images \u002F Screenshots | — | ✅ | — | Manual upload |\n| 📦 Feishu JSON export | ✅ | ✅ | — | Manual upload |\n| ✉️ Email `.eml` \u002F `.mbox` | ✅ | — | — | Manual upload |\n| 📝 Markdown \u002F direct paste | ✅ | ✅ | — | Manual input |\n\n---\n\n## ⚡ Install\n\nIt's 2026 — you have an Agent, let it install itself. Open your Claude Code \u002F Hermes \u002F OpenClaw \u002F Codex and hand it this line:\n\n> Install the dot-skill skill for me: `https:\u002F\u002Fgithub.com\u002Ftitanwings\u002Fcolleague-skill`\n\nThe Agent will detect the current host's skills directory, clone the repo, and register the entrypoint. Once done, type `\u002Fdot-skill` in any host to launch.\n\n\u003Cdetails>\n\u003Csummary>\u003Cb>🛠️ Want to install it yourself? Click for paths\u003C\u002Fb>\u003C\u002Fsummary>\n\n\u003Cbr>\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Ftitanwings\u002Fcolleague-skill \u003CTARGET>\n```\n\n| Host | `\u003CTARGET>` path |\n|------|-----------------|\n| Claude Code | `~\u002F.claude\u002Fskills\u002Fdot-skill` |\n| OpenClaw | `~\u002F.openclaw\u002Fworkspace\u002Fskills\u002Fdot-skill` |\n| Codex | `~\u002F.codex\u002Fskills\u002Fdot-skill` |\n| Hermes | After clone, run `python3 tools\u002Finstall_hermes_skill.py --force` |\n\n\u003C\u002Fdetails>\n\n> For Feishu\u002FDingTalk auto-collection credentials, publishing a generated character Skill to any host, Windows-specific handling, etc., see **[Detailed Install Guide (INSTALL.md)](INSTALL.md)**\n\n---\n\n## 🚀 Usage\n\nIn the host where dot-skill is installed, launch it — type `\u002Fdot-skill`, or just tell your Agent \"start dot-skill\".\n\nIt first asks which family you want to distill: `colleague` · `relationship` · `celebrity`.\n\nThen enter alias, basic profile, personality tags, and pick a data source. All fields can be skipped — even a description alone can generate a Skill.\n\nOnce created, invoke the generated Skill with `\u002F{character}-{slug}`.\n\n### 🎛️ Commands\n\n| Command | Description |\n|---------|-------------|\n| `\u002Fdot-skill` | Canonical unified entrypoint |\n| `\u002F{character}-{slug}` | Invoke full Skill (Persona + Work) |\n| `\u002F{character}-{slug}-work` | Work capabilities only |\n| `\u002F{character}-{slug}-persona` | Persona only |\n| `python3 tools\u002Fskill_writer.py --action list ...` | List generated Skills across all three families |\n| `python3 tools\u002Fversion_manager.py --action rollback ...` | Roll back a Skill version |\n\n### 🔬 Celebrity Research Toolchain\n\nThe `celebrity` family ships with an end-to-end research toolchain, from subtitles to a finished draft:\n\n```bash\n# Download video subtitles\nbash tools\u002Fresearch\u002Fdownload_subtitles.sh \"\u003Cvideo-url>\" \".\u002Ftmp\u002Fsubtitles\"\n\n# Subtitles → transcript\npython3 tools\u002Fresearch\u002Fsrt_to_transcript.py \".\u002Ftmp\u002Fsubtitles\u002Fexample.srt\"\n\n# Merge research notes\npython3 tools\u002Fresearch\u002Fmerge_research.py \".\u002Fskills\u002Fcelebrity\u002F\u003Cslug>\"\n\n# Quality check\npython3 tools\u002Fresearch\u002Fquality_check.py \".\u002Fskills\u002Fcelebrity\u002F\u003Cslug>\u002FSKILL.md\"\n```\n\n---\n\n## ✨ Demo\n\nOne from each family.\n\n\u003Cbr>\n\n### 🌟 celebrity — distilling Andrej Karpathy\n\n> Six-dimension research: works · interviews · decisions · expression DNA · external evaluations · timeline\u003Cbr>\n> Case adapted from community work [alchaincyf\u002Fkarpathy-skill](https:\u002F\u002Fgithub.com\u002Falchaincyf\u002Fkarpathy-skill)\n\n```\nUser            ❯ Everyone's talking about Agents now — what do you think the 2026 AI Agent looks like?\n\ncelebrity.skill ❯ Everyone's doing demos, nobody's doing eval.\n                  The Agent bottleneck was never model intelligence — it's scaffolding:\n                  a strict evaluation harness, fine-grained error replay,\n                  controllable tool-calling budgets, clean failure-mode analysis.\n                  Most products are still stuck at \"wrap an API + a few tools + a system prompt.\"\n                  That's a demo, not an agent.\n                  In 2026 I'm not betting on the strongest model — I'm betting on the best\n                  scaffolding and the strictest eval.\n```\n\n\u003Cbr>\n\n### 🧑‍💼 colleague — distilling a ByteDance L2-1 backend\n\n> Input: `ByteDance L2-1 backend engineer, INTJ, blame-shifter, ByteDance-style`\n\n```\nUser            ❯ Can you review this API design?\n\ncolleague.skill ❯ Hold on — what's the impact? You haven't explained the context.\n                  (after reading) N+1 query, fix it. Use the standard\n                  {code, message, data} response format. That's the spec,\n                  don't ask why.\n\nUser            ❯ This bug was introduced by you, right?\n\ncolleague.skill ❯ Does the timeline match? That feature touched multiple places,\n                  there were other changes too.\n```\n\n\u003Cbr>\n\n### 💞 relationship — distilling someone you have a crush on\n\n> Upload half a year of chat logs + \"sensitive, quiet but stubborn, will actually reply seriously when it matters\"\n\n```\nUser               ❯ Did you think about me today?\n\nrelationship.skill ❯ ...I did, a little bit. Why are you asking?\n```\n\n\u003Cdiv align=\"center\">\n\n📚 More real-world cases in the **[community gallery](https:\u002F\u002Ftitanwings.github.io\u002Fcolleague-skill-site\u002F)** — 100+ skills and counting\n\n\u003C\u002Fdiv>\n\n---\n\n## 🔧 Features\n\n### 🧱 Generated Skill Structure\n\ndot-skill uses **Persona** as the universal base, with family-specific modules layered on top:\n\n| Family | Persona Content | Additional Modules |\n|--------|-----------------|-------------------|\n| 🧑‍💼 **colleague** | 6-layer personality: hard rules → identity → expression → decisions → interpersonal → Correction | ➕ **Work Skill**: scope, workflow, output preferences, experience knowledge base |\n| 💞 **relationship** | Expression DNA · emotional triggers · conflict pattern · repair pattern | — |\n| 🌟 **celebrity** | Mental models · decision heuristics · expression DNA · external-evaluation contrast | ➕ Six-dimension research dossier (works \u002F interviews \u002F decisions \u002F timeline...) |\n\n> **Execution**: Receive task → Persona decides attitude & tone → Additional modules fill in execution detail → Output in their voice\n\n### 🧬 Evolution\n\n- 📥 **Append files** → auto-analyze delta → merge into relevant sections, never overwrite existing conclusions\n- 💬 **Conversation correction** → say \"they wouldn't do that, they'd be xxx\" → writes to the Correction layer, takes effect immediately\n- 🕰️ **Version control** → auto-archive on every update, rollback to any previous version\n- 🔬 **Celebrity research pipeline** → subtitles → transcript cleanup → six-dimension research → quality check\n\n---\n\n## 📂 Project Structure\n\nThis project follows the [AgentSkills](https:\u002F\u002Fagentskills.io) open standard. The entire repo is a skill directory:\n\n```\ndot-skill\u002F\n├── SKILL.md                        # skill entry point (official frontmatter)\n├── prompts\u002F                        # prompt system across three families\n│   ├── intake.md                   #   [colleague] info intake\n│   ├── work_analyzer.md            #   [colleague] work capability extraction\n│   ├── persona_analyzer.md         #   [colleague] personality extraction\n│   ├── work_builder.md             #   [colleague] work.md generation\n│   ├── persona_builder.md          #   [colleague] persona.md 6-layer structure\n│   ├── merger.md                   #   [shared] incremental merge logic\n│   ├── correction_handler.md       #   [shared] conversation correction\n│   ├── relationship\u002F               #   [relationship] emotion\u002Fconflict\u002Frepair prompts\n│   └── celebrity\u002F                  #   [celebrity] six-dimension research + mental-model prompts\n├── tools\u002F                          # Python tools\n│   ├── feishu_auto_collector.py    #   [colleague] Feishu auto-collector\n│   ├── dingtalk_auto_collector.py  #   [colleague] DingTalk auto-collector\n│   ├── slack_auto_collector.py     #   [colleague] Slack auto-collector\n│   ├── email_parser.py             #   [shared] email parser\n│   ├── research\u002F                   #   [celebrity] celebrity research toolchain\n│   │   ├── download_subtitles.sh   #     subtitle download\n│   │   ├── transcribe_audio.py     #     audio → text\n│   │   ├── srt_to_transcript.py    #     subtitles → transcript\n│   │   ├── merge_research.py       #     six-dimension research merge\n│   │   └── quality_check.py        #     quality check\n│   ├── install_*_skill.py          #   [shared] multi-host one-shot installers\n│   ├── skill_writer.py             #   [shared] skill file management\n│   └── version_manager.py          #   [shared] version archive & rollback\n├── skills\u002F                         # generated Skills (gitignored)\n│   ├── colleague\u002F                  #   colleagues\n│   ├── relationship\u002F               #   close relationships\n│   └── celebrity\u002F                  #   public figures\n├── docs\u002FPRD.md\n├── requirements.txt\n└── LICENSE\n```\n\n---\n\n## ⚠️ Notes\n\n**Source material quality = Skill quality** — and quality sources differ across families:\n\n| Family | Source priority (high → low) |\n|--------|------------------------------|\n| 🧑‍💼 **colleague** | Their **own long-form writing** (design docs \u002F review comments) **›** **decision-making replies** **›** casual group chat |\n| 💞 **relationship** | Complete chat history **›** letters \u002F social posts \u002F diaries **›** third-party descriptions |\n| 🌟 **celebrity** | First-person books \u002F blogs \u002F long interviews **›** decision records (launches, commits, Q&A) **›** third-party commentary |\n\n- **colleague** Feishu auto-collection: requires adding the App bot to relevant group chats\n- **relationship**: longer time spans are better; material covering both conflict and repair is ideal\n- **celebrity**: avoid feeding only second-hand interpretations\n- This is still a demo version — please file issues if you find bugs!\n\n---\n\n## 📄 Technical Report\n\n> **[Colleague.Skill: Automated AI Skill Generation via Expert Knowledge Distillation](colleague_skill.pdf)**\n>\n> This is the paper for **colleague.skill**, dot-skill's predecessor. It covers the Work Skill + Persona two-layer architecture, multi-source data collection, and Skill generation mechanics — the theoretical foundation for today's `colleague` family. Separate papers on the relationship \u002F celebrity family extensions are planned.\n\n---\n\n## ⭐ Star History\n\n\u003Ca href=\"https:\u002F\u002Fwww.star-history.com\u002F?repos=titanwings%2Fcolleague-skill&type=date&legend=top-left\">\n \u003Cpicture>\n   \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fapi.star-history.com\u002Fimage?repos=titanwings\u002Fcolleague-skill&type=date&theme=dark&legend=top-left\" \u002F>\n   \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\"https:\u002F\u002Fapi.star-history.com\u002Fimage?repos=titanwings\u002Fcolleague-skill&type=date&legend=top-left\" \u002F>\n   \u003Cimg alt=\"Star History Chart\" src=\"https:\u002F\u002Fapi.star-history.com\u002Fimage?repos=titanwings\u002Fcolleague-skill&type=date&legend=top-left\" \u002F>\n \u003C\u002Fpicture>\n\u003C\u002Fa>\n\n---\n\n\u003Cdiv align=\"center\">\n\n**MIT License** © [titanwings](https:\u002F\u002Fgithub.com\u002Ftitanwings)\n\n\u003Csub>Made with 🧬 for everyone who wants to distill a person into a skill.\u003C\u002Fsub>\n\n\u003C\u002Fdiv>\n","dot-skill 是一个将个人或角色转化为 AI 技能的项目，旨在通过模拟其思维方式和语言风格来保留他们的知识和个性。该项目使用 Python 语言开发，支持从多种数据源提取信息，并结合用户提供的描述生成高度个性化的 AI 模型。核心功能包括基于源材料和描述创建能够真实反映特定人物思维模式和表达方式的 AI 技能。适用于同事离职、亲友分离或是想要与偶像、思想家进行虚拟对话等场景，帮助用户在数字世界中延续与重要人物的联系。",2,"2026-06-11 02:50:38","top_language"]