[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-76054":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":10,"languages":10,"totalLinesOfCode":10,"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":14,"compositeScore":18,"rankGlobal":10,"rankLanguage":10,"license":19,"archived":20,"fork":20,"defaultBranch":21,"hasWiki":22,"hasPages":20,"topics":23,"createdAt":10,"pushedAt":10,"updatedAt":24,"readmeContent":25,"aiSummary":26,"trendingCount":15,"starSnapshotCount":15,"syncStatus":27,"lastSyncTime":28,"discoverSource":29},76054,"yansu-skill","Isoform\u002Fyansu-skill","Isoform","Bespoke context, delivered to your AI agent. Yansu observes how you actually work and crystallizes it into knowledge — this skill is how that knowledge reaches whichever agent you happen to be using. Prompt once, never re-introduce yourself.","https:\u002F\u002Fyansu.app",null,157,5,138,1,0,6,20,2.33,"MIT License",false,"main",true,[],"2026-06-12 02:03:39","\u003Cdiv align=\"center\">\n\n# Yansu.skill\n\n> *\"Other agents start cold. Yansu.skill starts warm.\"*\n\n[![License: MIT](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-yellow.svg)](LICENSE)\n[![Yansu](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FYansu-Skill-d4603a)](https:\u002F\u002Fyansu.app)\n\n[English](README.md) · [中文](README_ZH.md)\n\n\u003Cbr>\n\n**Bespoke context, delivered to your agent. Prompt once, never re-introduce yourself.**\n\n[Yansu](https:\u002F\u002Fyansu.app) listens, crystallizes, and quietly builds. **Yansu.skill** is how that crystallized knowledge reaches whichever agent you happen to be using — so it walks in already knowing how you actually work.\n\n\u003Cbr>\n\n**Install** — paste this to your AI agent:\n\n\u003C\u002Fdiv>\n\n```\nInstall the Yansu skill for me:\n\n1. Fetch https:\u002F\u002Fraw.githubusercontent.com\u002FIsoform\u002Fyansu-skill\u002Fmain\u002FSKILL.md\n2. Save it under your own user-level skills directory as `yansu\u002FSKILL.md`\n   (whatever path your agent reads skills from on this machine).\n3. Confirm the install path when done.\n```\n\n\u003Cdiv align=\"center\">\n\n[See it in action](#see-it-in-action) · [Features](#features) · [How it works](#how-it-works) · [Privacy](#privacy)\n\n\u003C\u002Fdiv>\n\n---\n\n## See it in action\n\nType `\u002Fyansu` to invoke the skill explicitly, or ask naturally and your agent will pull it in when it spots a continuity question.\n\n```\nYou      ❯ \u002Fyansu Why did that build fail again — the auth one?\n\nAgent    ❯ Same migration-lock issue from a few weeks back. You\n           worked around it last time by running the ALTER inside\n           a transaction with statement_timeout off. Want me to\n           apply that recipe?\n```\n\n```\nYou      ❯ \u002Fyansu Spin up the staging database\n\nAgent    ❯ You've done this twice. Your compose template lives\n           next door, and the first boot usually trips on the SSL\n           cert path. Starting from your last-known-good and\n           watching for the cert thing.\n```\n\n```\nYou      ❯ \u002Fyansu Pick the right tool for the job\n\nAgent    ❯ For this kind of task you usually reach for the same\n           three. The first one fits the shape of the problem,\n           and you've been getting cleaner output out of it since\n           you tuned the flags last week.\n```\n\nThis is not role-play. This is continuity.\n\n---\n\n## Features\n\n### Tribal Knowledge\n\nYour agent inherits everything Yansu has crystallized — every conversation, decision, hesitation, and pattern. Already structured, already tagged, already yours. The knowledge that used to live only in your head now lives wherever your agent does.\n\n### No Re-introduction\n\nYou still prompt your agent — you just stop re-explaining yourself. The agent already knows your tools, your workflows, the gotchas you've paid for. Say *\"help me ship the thing\"* and it knows what the thing is. Say *\"the bug from last week\"* and it knows which bug.\n\n### Distraction Free\n\nYansu.skill stays silent until your question actually needs personal context. No noisy preambles. No agent pretending to remember things it doesn't. Continuity when you need it, invisible when you don't.\n\n---\n\n## How it works\n\n**1. Yansu listens.** It observes how you work — your desktop, your messaging apps, the conversations and decisions that would otherwise vanish.\n\n**2. Yansu crystallizes.** Raw memory becomes structured, actionable knowledge — patterns you can reuse, not transcripts you'd never re-read.\n\n**3. Yansu.skill surfaces.** When your agent needs you, the skill hands over the relevant slice — one entry, one sentence, never the whole archive.\n\nYansu does the listening. The skill does the handoff.\n\n---\n\n## Privacy\n\n**Lives on your machine.** Your memory, your captured workflows, your generated knowledge — all stored locally. Not on our servers. On yours.\n\n**You hold the keys.** Nothing leaves your device through this skill. Delete an entry to make your agent forget it; uninstall the skill to disable the connection entirely.\n\n**Read-only by design.** Capture is Yansu's job. Recall is the skill's. The skill will never write back to your memory, your knowledge, or your machine.\n\n---\n\n## Honest boundaries\n\nA skill that won't tell you what it can't do is one you can't trust.\n\n- **It doesn't predict the future** — only your past, as captured by Yansu.\n- **It only knows what Yansu has seen** — if you've never used Yansu while doing X, your agent won't know about X.\n- **It is quiet by default** — it stays out of every turn that doesn't need it.\n\nA continuity layer that pretends to be omniscient isn't continuity. It is flattery.\n\n---\n\n## License\n\nMIT — see [LICENSE](LICENSE).\n\nBuilt for [Yansu](https:\u002F\u002Fyansu.app) by [Isoform](https:\u002F\u002Fgithub.com\u002FIsoform).\n","Yansu.skill 项目旨在为你的AI助手提供定制化的上下文信息，使得助手能够基于你实际工作中的知识和习惯进行更有效的交互。其核心功能在于通过观察并提炼用户的工作模式与偏好，将这些个性化知识无缝集成到任何正在使用的AI代理中，实现“一次设置，永久记忆”的效果。技术特点包括自动继承所有已记录的对话、决策及行为模式，无需重复介绍个人偏好或历史背景，从而在需要时直接提供相关帮助。适用于希望提高工作效率、减少重复沟通成本，并且希望通过AI助手获得更加个性化的支持场景。",2,"2026-06-11 03:54:19","CREATED_QUERY"]