SKILL

ynh-adopt

Bring an existing AI setup — a .claude folder, .cursorrules, a grown CLAUDE.md — under ynh without losing it. Use when someone already has assistant config and wants to adopt ynh, rather than starting a harness from scratch.

原文 · 译文尚未补齐

适用对象:对象待分类。自动收录 · 待核查。

原作者说明节选

从固定版本的 SKILL.md 提取,保留原文语言。以下内容属于作者说明,本站尚未验证安装、兼容或运行结果。

查看英文原文

原文涵盖的主题

  • Adopt an existing setup
  • Before you start
  • Step 1 — Inventory, out loud
  • Step 2 — Deal with the instructions file first
  • Step 3 — Pick a strategy
  • A. Include in place — config stays in this repo
  • B. Move into a harness — config becomes reusable
  • Step 4 — Legacy manifest

Almost nobody starts from nothing. They arrive with a .claude/ folder someone added six months ago, a .cursorrules from before that, and a CLAUDE.md that has grown by accretion. The question is never "how do I build a harness". It is "what happens to what I already have?" Answer that first, or nothing else lands. If they genuinely have nothing, use ynh-create-harness instead. references/what-moves-where.md — the decision table, the traps, and the two adoption strategies in full Look, and tell them what you found. Do not guess. What tends to be there: | Found | What it is | |---|---| | .claude/skills/, .claude/agents/, .claude/commands/ | already in the artifact format ynh uses | | .claude/settings.json | permissions and hooks — not an artifact, stays put | | CLAUDE.md | project instructions, Claude-specific filename | | .cursorrules | project instructions, Cursor's legacy filename | | .cursor/rules/.mdc | Cursor rules with frontmatter | | .github/copilot-instructions.md | Copilot instructions | | AGENTS.md | already the cross-vendor form — the easy case | | .harness.json | a legacy ynh manifest — run ynd migrate | Then ask the question that decides everything else: Do you want this

这里只展示节选,请以完整原文和当前要求为准。

来源仓库
eyelock/ynh
源文件
skills/ynh-adopt/SKILL.md
固定版本
3a2944a97f6341a43490dac4e5b1fd37e8c3b6e6
阅读完整原作者说明 ↗

准备条件

  • 本条仅收录原作者的技能元数据;请阅读完整说明,确认依赖、许可与使用条件。

第一件可以尝试的小事

先阅读原作者说明,再决定是否在自己的 AI 工具中使用。

对应 AI 工具说明

作者有说明不等于本站实机验证。没有证据时,不自动标为兼容、安全或免费。

来源与内容状态

本条为自动发现的元数据,尚未完成人工文档核查;适用性与运行结果待核查。

原作者:eyelock/ynh · 原作者仓库元数据。文档核查:未核查。最近来源检查:2026-11-10T23:01:52.100Z。

分类由名称、说明与路径规则生成,分类不代表已核实适配。

许可与权利归属

MIT。第三方内容及版权归原作者,使用条件以原文为准。

许可原文 ↗

本站没有因同步、复制或用户自报成功而将资源标为运行验证通过。

限制与注意事项

  • 元数据收录不代表本站已经安装、适配、实测或进行安全审查。