[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M72ADHX7EXZ7KV9FG1MEEVG8":3,"skill-same:01M72ADHX7EXZ7KV9FG1MEEVG8":59},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":23,"commitSha":24,"skillDirectory":25,"forksCount":26,"repositoryHtmlUrl":27,"repositoryHomepageUrl":28,"ownerHtmlUrl":29,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":30,"install":31,"resources":44},"01M72ADHX7EXZ7KV9FG1MEEVG8","using-xskill",null,"Use when installing, configuring, or operating xskill (the `xskill` CLI \u002F `pip install xskill`) — starting the daemon, registering trajectory dirs, joining a team server, understanding how trajectories become Skills, or rebuilding\u002Fre-distilling the skill library after a model change.","SkillNerds\u002Fxskill","SkillNerds","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F283534061?v=4","01M72ADC90GC2G52CXN8DKGP5Q","01M72ADC9KGMTQHYDEGS3FG98H",129,"MIT","https:\u002F\u002Fgithub.com\u002FSkillNerds\u002Fxskill\u002Fblob\u002Fmain\u002Fskills\u002Fusing-xskill\u002FSKILL.md",[],[],"2026-11-10 18:39:20",false,"---\nname: using-xskill\ndescription: Use when installing, configuring, or operating xskill (the `xskill` CLI \u002F `pip install xskill`) — starting the daemon, registering trajectory dirs, joining a team server, understanding how trajectories become Skills, or rebuilding\u002Fre-distilling the skill library after a model change.\n---\n\n# Using xskill\n\n## Overview\n\nxskill distills reusable **Skills** (`SKILL.md` folders) out of the real execution\ntrajectories of coding agents (Claude Code, Codex, OpenCode, Cursor, …). A background\ndaemon watches each agent's session logs, slices them into single-intent **atoms**,\nclusters atoms into skills, and writes\u002Fversions each skill in its own git folder. New\nskill versions only replace old ones when real traffic shows they serve users better\n(canary A\u002FB by UX score) — not by an LLM grading itself.\n\n**Core mental model:** raw trajectory → atoms → candidate routing → SKILL.md → canary\nA\u002FB → installed into every agent's skill dir. You operate the daemon; the daemon does\nthe distilling.\n\n## When to Use\n\n- Installing xskill or filling in `~\u002F.xskill\u002Fconfig.yaml` (LLM + embedding endpoints)\n- Starting\u002Fkeeping the daemon running (`xskill serve`), or backfilling old trajectories\n- Joining or hosting a team server (`xskill serve --server` \u002F `xskill connect`)\n- Understanding the agent pipeline, atoms, canary\u002FUX scoring, or deployment modes\n- **Re-distilling the whole skill library** (e.g. after switching to a stronger model)\n\n## Quick Reference\n\n| Command | What it does |\n|---------|--------------|\n| `pip install xskill` | Install (Python 3.9+) |\n| `xskill serve` | Standalone daemon: FastAPI + watcher; first run writes `~\u002F.xskill\u002Fconfig.yaml` then exits |\n| `xskill serve --server` | Team server: owns all LLM calls + git; prints a join token |\n| `xskill connect \u003Chost:port> --token \u003Ct>` | Join a team server as a thin client |\n| `xskill registry add \u003Cpath>` | Backfill \u002F watch an extra trajectory directory |\n| `xskill traj search \u003Cquery>` \u002F `xskill search \u003Cquery>` | Search trajectories or skills |\n| `xskill read \u003Cpath> --eco \u003Ceco>` | Batch-ingest db trajectories (ngagent\u002Fopencode) |\n| `xskill rebuild [--force]` | Re-distill from existing raw trajectories (see reference) |\n| `xskill stats` | Token usage & estimated cost |\n\nThe daemon is the engine: most commands only change state in the DB; nothing is\ndistilled unless `xskill serve` (or the team server) is running.\n\n## Progressive Disclosure — read on demand\n\n- **Install & configure** (config.yaml fields, per-agent collect\u002Finstall paths, team\n  client setup): `references\u002Finstallation.md`\n- **How it works** (TaskAgent → TaskClusterAgent → SkillEditAgent, atoms, canary\u002FUX\n  scoring, standalone vs team mode): `references\u002Fmechanisms.md`\n- **Rebuild the skill library** (a ready-to-run prompt that walks a model through\n  re-distilling correctly): `references\u002Frebuilding-skill-library.md`\n\n## Common Mistakes\n\n- **Running `rebuild` with no daemon up.** `rebuild` only resets DB state; the watcher\n  in `serve` does the actual re-split\u002Fre-cluster every 30s. No daemon = nothing happens.\n- **Deleting raw `~\u002F.xskill\u002F*_sessions\u002F*.md`.** Those are the *input* to distillation —\n  delete them and you can no longer rebuild.\n- **Expecting DeepSeek to do embeddings.** DeepSeek has no embedding endpoint; point the\n  `embedding:` block at DashScope \u002F OpenAI \u002F Ollama.\n- **Putting tokens in public places.** Team join tokens must never land in a public repo\n  or chat log.\n","# Using xskill\n\n## Overview\n\nxskill distills reusable **Skills** (`SKILL.md` folders) out of the real execution\ntrajectories of coding agents (Claude Code, Codex, OpenCode, Cursor, …). A background\ndaemon watches each agent's session logs, slices them into single-intent **atoms**,\nclusters atoms into skills, and writes\u002Fversions each skill in its own git folder. New\nskill versions only replace old ones when real traffic shows they serve users better\n(canary A\u002FB by UX score) — not by an LLM grading itself.\n\n**Core mental model:** raw trajectory → atoms → candidate routing → SKILL.md → canary\nA\u002FB → installed into every agent's skill dir. You operate the daemon; the daemon does\nthe distilling.\n\n## When to Use\n\n- Installing xskill or filling in `~\u002F.xskill\u002Fconfig.yaml` (LLM + embedding endpoints)\n- Starting\u002Fkeeping the daemon running (`xskill serve`), or backfilling old trajectories\n- Joining or hosting a team server (`xskill serve --server` \u002F `xskill connect`)\n- Understanding the agent pipeline, atoms, canary\u002FUX scoring, or deployment modes\n- **Re-distilling the whole skill library** (e.g. after switching to a stronger model)\n\n## Quick Reference\n\n| Command | What it does |\n|---------|--------------|\n| `pip install xskill` | Install (Python 3.9+) |\n| `xskill serve` | Standalone daemon: FastAPI + watcher; first run writes `~\u002F.xskill\u002Fconfig.yaml` then exits |\n| `xskill serve --server` | Team server: owns all LLM calls + git; prints a join token |\n| `xskill connect \u003Chost:port> --token \u003Ct>` | Join a team server as a thin client |\n| `xskill registry add \u003Cpath>` | Backfill \u002F watch an extra trajectory directory |\n| `xskill traj search \u003Cquery>` \u002F `xskill search \u003Cquery>` | Search trajectories or skills |\n| `xskill read \u003Cpath> --eco \u003Ceco>` | Batch-ingest db trajectories (ngagent\u002Fopencode) |\n| `xskill rebuild [--force]` | Re-distill from existing raw trajectories (see reference) |\n| `xskill stats` | Token usage & estimated cost |\n\nThe daemon is the engine: most commands only change state in the DB; nothing is\ndistilled unless `xskill serve` (or the team server) is running.\n\n## Progressive Disclosure — read on demand\n\n- **Install & configure** (config.yaml fields, per-agent collect\u002Finstall paths, team\n  client setup): `references\u002Finstallation.md`\n- **How it works** (TaskAgent → TaskClusterAgent → SkillEditAgent, atoms, canary\u002FUX\n  scoring, standalone vs team mode): `references\u002Fmechanisms.md`\n- **Rebuild the skill library** (a ready-to-run prompt that walks a model through\n  re-distilling correctly): `references\u002Frebuilding-skill-library.md`\n\n## Common Mistakes\n\n- **Running `rebuild` with no daemon up.** `rebuild` only resets DB state; the watcher\n  in `serve` does the actual re-split\u002Fre-cluster every 30s. No daemon = nothing happens.\n- **Deleting raw `~\u002F.xskill\u002F*_sessions\u002F*.md`.** Those are the *input* to distillation —\n  delete them and you can no longer rebuild.\n- **Expecting DeepSeek to do embeddings.** DeepSeek has no embedding endpoint; point the\n  `embedding:` block at DashScope \u002F OpenAI \u002F Ollama.\n- **Putting tokens in public places.** Team join tokens must never land in a public repo\n  or chat log.\n",{"name":5,"description":7},"main","d438329390736e6ea696c4fe25c53a72a06eda37","skills\u002Fusing-xskill",27,"https:\u002F\u002Fgithub.com\u002FSkillNerds\u002Fxskill","https:\u002F\u002Fxskill.wiki\u002F","https:\u002F\u002Fgithub.com\u002FSkillNerds","2026-11-11 18:14:29",{"source":8,"skillName":5,"commands":32,"prompt":42,"zipAvailable":43},[33,36,39],{"tool":34,"command":35},"npx","npx skills add SkillNerds\u002Fxskill --skill using-xskill",{"tool":37,"command":38},"bunx","bunx skills add SkillNerds\u002Fxskill --skill using-xskill",{"tool":40,"command":41},"pnpm","pnpm dlx skills add SkillNerds\u002Fxskill --skill using-xskill","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FSkillNerds\u002Fxskill\u002Fblob\u002Fmain\u002Fskills\u002Fusing-xskill\u002FSKILL.md",true,[45,51,55],{"path":46,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":49,"sourceHtmlUrl":50},"references\u002Finstallation.md","REFERENCE","参考文档",3865,"https:\u002F\u002Fgithub.com\u002FSkillNerds\u002Fxskill\u002Fblob\u002Fmain\u002Fskills\u002Fusing-xskill\u002Freferences\u002Finstallation.md",{"path":52,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":53,"sourceHtmlUrl":54},"references\u002Fmechanisms.md",3781,"https:\u002F\u002Fgithub.com\u002FSkillNerds\u002Fxskill\u002Fblob\u002Fmain\u002Fskills\u002Fusing-xskill\u002Freferences\u002Fmechanisms.md",{"path":56,"resourceType":47,"resourceTypeDesc":48,"sizeBytes":57,"sourceHtmlUrl":58},"references\u002Frebuilding-skill-library.md",5165,"https:\u002F\u002Fgithub.com\u002FSkillNerds\u002Fxskill\u002Fblob\u002Fmain\u002Fskills\u002Fusing-xskill\u002Freferences\u002Frebuilding-skill-library.md",[60,67,74,81,88],{"skillNo":61,"name":62,"nameZh":6,"description":63,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":64,"categories":65,"keywords":66,"publishedAt":18},"01M72ADW9S00NVWZSWDNW5VZ90","demo-skill","A demo skill used to verify openclaw startup install path","https:\u002F\u002Fgithub.com\u002FSkillNerds\u002Fxskill\u002Fblob\u002Fmain\u002Ftests\u002Fdocker_e2e\u002Fscenarios\u002Fruntime_openclaw_detect\u002Fpre_state\u002F.xskill\u002Fskill\u002Fdemo-skill\u002FSKILL.md",[],[],{"skillNo":68,"name":69,"nameZh":6,"description":70,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":71,"categories":72,"keywords":73,"publishedAt":18},"01M72ADV32DN09D8FFD049799J","xskill-demo-marker","A unique marker skill installed by xskill, used to verify openclaw sees it","https:\u002F\u002Fgithub.com\u002FSkillNerds\u002Fxskill\u002Fblob\u002Fmain\u002Ftests\u002Fdocker_e2e\u002Fopenclaw_real_llm\u002Fdemo_skill\u002FSKILL.md",[],[],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":80,"publishedAt":18},"01M72ADQDB89DNXR7QB6YR5YTC","xskill-helper","Use this skill for xskill. Trigger when the user wants to init, connect or join a team server, generate or rewrite a team Skill from trajectories, search download upload or import team Skills, look up this user or a teammate's past chats across Claude Code Codex Cursor OpenCode Trae DeepSeek Harness and ngagent, upgrade the client, or debug a stuck daemon. Team chats: xskill traj search then traj read. This machine only: add --local to read ~\u002F.xskill\u002F*_sessions. Invoke as \u002Fxskill-helper.","https:\u002F\u002Fgithub.com\u002FSkillNerds\u002Fxskill\u002Fblob\u002Fmain\u002Fsrc\u002Fxskill\u002Fdata\u002Fskill\u002Fxskill-helper\u002FSKILL.md",[],[],{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":87,"publishedAt":18},"01M72ADMRHYGZ6Y73BPK82C60A","xskill-dev","Use when developing xskill itself — editing agent tool docstrings, changing the generate agent tool surface, or verifying what schema the Agno framework actually sends to the model. Covers the dump_schema workflow and where tool descriptions come from.","https:\u002F\u002Fgithub.com\u002FSkillNerds\u002Fxskill\u002Fblob\u002Fmain\u002Fskills\u002Fxskill-dev\u002FSKILL.md",[],[],{"skillNo":89,"name":90,"nameZh":6,"description":91,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":92,"categories":93,"keywords":94,"publishedAt":18},"01M72ADFZV2TBW575YCCKWJDMG","xskill-registry-db-first","XSkill 数据访问原则：文件系统是内容真相，由 worker 同步进库；面板与业务读 请求只走数据库；内核 agent loop 可直接写盘。适用于改 dashboard、recommend、 team、pipeline、ecosystems 等读路径，以及审查是否又引入业务侧扫盘。","https:\u002F\u002Fgithub.com\u002FSkillNerds\u002Fxskill\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fxskill-registry-db-first\u002FSKILL.md",[],[]]