[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M78EXG55D8CSHZR9V4HBAP76":3,"skill-same:01M78EXG55D8CSHZR9V4HBAP76":65},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":6,"ownerHtmlUrl":29,"ownerBlogUrl":30,"ownerDescription":31,"updatedAt":32,"install":33,"resources":46},"01M78EXG55D8CSHZR9V4HBAP76","adapters",null,"Configure and verify ContextOS exports for Gemini\u002FAntigravity, Claude Code, Cursor, GitHub Copilot, Aider, and Zed without editing generated projections.","kok-o\u002Fcontextos-agents","kok-o","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F173567992?v=4","01M78EV59TFZZ32P2FNN27BHW5","Nurkhan Esenbek","01M78EV5AFG9E8PEC10N3W7JNH",10,"Apache-2.0","https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Fcatalog\u002Fskills\u002Fadapters\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: adapters\ndescription: >\n  Configure and verify ContextOS exports for Gemini\u002FAntigravity, Claude Code,\n  Cursor, GitHub Copilot, Aider, and Zed without editing generated projections.\n---\n\n# adapters\n\n## Overview\n\nExport installed ContextOS rules into six supported client formats. This skill\ndocuments the exporters; the CLI performs compilation and filesystem changes.\nGenerated files prove delivery, not client loading or model adherence.\n\n## When to Use\n\nConfigure, synchronize, or diagnose agent rules across the supported clients.\nDo not assume a catalog skill is installed just because it exists in the catalog.\n\n## Rules & Patterns\n\nInspect the target project, winning manifest, declared entrypoint, resources,\nand active profile first. Canonical built-ins live in `.agents\u002Fcore\u002Fskills\u002F`;\nproject additions and overrides belong in `.agents\u002Fproject\u002Fskills\u002F`. Plugins\nprovide additional sources. Never edit native projections to change source rules.\n\n## Supported Agents\n\nThe table shows the default `auto` outputs. Replace `\u003Cskill>` with the installed\nskill ID. Declared resources are also copied beside Gemini and Claude auto skills.\n\n| Agent | Config Target | Delivery |\n| --- | --- | --- |\n| **Gemini \u002F Antigravity** | `.agents\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Fgenerated\u002Fgemini\u002Fskills\u002F\u003Cskill>\u002FSKILL.md` | Shared native skill and generated copy |\n| **Claude Code** | `.claude\u002Fskills\u002F\u003Cskill>\u002FSKILL.md` | Native on-demand skill |\n| **GitHub Copilot** | `.github\u002Fcopilot-instructions.md` | Managed block with source paths and routing descriptions |\n| **Cursor** | `.cursor\u002Frules\u002F\u003Cskill>.mdc` | MDC rule with description, globs, and alwaysApply |\n| **Aider** | `.aider.conf.yml`, `CONVENTIONS.md` | YAML read entry for conventions; on-demand source references |\n| **Zed IDE** | `.zed\u002Frules.md`, `.zed\u002Fprompts\u002F\u003Cskill>.md` | Rule index and separate prompt body |\n\nWhen `.agents\u002FAGENTS.md` exists, Cursor also emits\n`.cursor\u002Frules\u002F00-project-rules.mdc`, Zed emits `.zed\u002Fprompts\u002Fcontextos.md`,\nand Copilot, Aider, and Zed include project rules in their instruction files.\nClaude does not generate `CLAUDE.md` or `.agents\u002Fgenerated\u002Fclaude\u002F`;\nCursor does not generate a legacy `.cursorrules` fallback.\n\n## Activation\n\nSet `activation` and `globs` in the winning manifest or entrypoint frontmatter.\nAbsent both, known catalog path scopes apply; otherwise the default is `auto`.\nThe resolver's `signals.fileGlobs` affects recommendations, not adapter activation.\n\n| Target | Activation | Native artifacts | Metadata \u002F behavior |\n| --- | --- | --- | --- |\n| `gemini` | `auto` | `.agents\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Fgenerated\u002Fgemini\u002Fskills\u002F\u003Cskill>\u002FSKILL.md` | Description-based discovery; no activation rule |\n| `gemini` | `always` | `.agents\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Fgenerated\u002Fgemini\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Frules\u002Fcontextos-\u003Cskill>.md` | Rule trigger is always_on |\n| `gemini` | `globs` | `.agents\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Fgenerated\u002Fgemini\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Frules\u002Fcontextos-\u003Cskill>.md` | Rule trigger is glob; additional bounded Windows rule files may be emitted |\n| `claude` | `auto` | `.claude\u002Fskills\u002F\u003Cskill>\u002FSKILL.md` | Name and description frontmatter; body on demand |\n| `claude` | `always` | `.claude\u002Frules\u002F\u003Cskill>.md` | Full body without paths frontmatter |\n| `claude` | `globs` | `.claude\u002Frules\u002F\u003Cskill>.md` | Full body with paths list |\n| `cursor` | `auto` | `.cursor\u002Frules\u002F\u003Cskill>.mdc` | Agent-requested description, empty globs, alwaysApply false |\n| `cursor` | `always` | `.cursor\u002Frules\u002F\u003Cskill>.mdc` | Empty globs, alwaysApply true |\n| `cursor` | `globs` | `.cursor\u002Frules\u002F\u003Cskill>.mdc` | Declared globs, alwaysApply false |\n\nCopilot, Aider, and Zed inline `always` skill bodies in their instruction files.\nOther skills remain routing references there; Zed also emits every skill's prompt\nbody. These three adapters do not enforce per-file glob activation. Markdown\nresource links are relocated when entrypoints move; load supporting files only\nwhen needed. Export selects installed skills allowed by the profile, not just the\nresolver's recommendation for the current task.\n\n## Generation Commands\n\n- `contextos compile` validates manifests and rebuilds the registry.\n- `contextos export \u003Cagent>` renders and applies one of the six targets above.\n- `contextos export all` applies all adapters in one transaction.\n- `contextos export all --check --json` checks drift without writing outputs.\n- `contextos export all --dry-run --json` inspects planned artifacts.\n\nIn this source checkout, use `node .agents\u002Fctx.js` instead of `contextos`.\nAfter source edits, compile, export, then check. A successful check means the\nselected projections match their sources and recorded ownership; it does not\nprove that a live client loaded them. Review ownership conflicts instead of\nautomatically forcing replacement. Aider merges its YAML read list and preserves\nexisting settings; inspect new defaults before running the client.\n\n## Code Examples\n\nSee [EXAMPLES.md](EXAMPLES.md) for the default adapters skill outputs and a\npath-scoped activation example.\n\n## Validation Checklist\n\n- [ ] Correct project root, winning manifest, entrypoint, resources, and profile.\n- [ ] Expected native path and activation metadata for each selected client.\n- [ ] Compile succeeds, export preserves user changes, and check reports no drift.\n- [ ] Generated-file evidence and any live-client testing are reported separately.\n\n## Common Mistakes\n\nEditing outputs, confusing resolver selection with export scope, assuming every\nadapter implements file globs, or treating a generated file as model-quality proof.\nDo not expect Continue or OpenHands exports; they are not registered targets.\n\n## Integration Notes\n\nUse context-os for manifest, resolver, profile, and budget configuration;\nengineering-workflow owns proportional implementation and verification.\n","# adapters\n\n## Overview\n\nExport installed ContextOS rules into six supported client formats. This skill\ndocuments the exporters; the CLI performs compilation and filesystem changes.\nGenerated files prove delivery, not client loading or model adherence.\n\n## When to Use\n\nConfigure, synchronize, or diagnose agent rules across the supported clients.\nDo not assume a catalog skill is installed just because it exists in the catalog.\n\n## Rules & Patterns\n\nInspect the target project, winning manifest, declared entrypoint, resources,\nand active profile first. Canonical built-ins live in `.agents\u002Fcore\u002Fskills\u002F`;\nproject additions and overrides belong in `.agents\u002Fproject\u002Fskills\u002F`. Plugins\nprovide additional sources. Never edit native projections to change source rules.\n\n## Supported Agents\n\nThe table shows the default `auto` outputs. Replace `\u003Cskill>` with the installed\nskill ID. Declared resources are also copied beside Gemini and Claude auto skills.\n\n| Agent | Config Target | Delivery |\n| --- | --- | --- |\n| **Gemini \u002F Antigravity** | `.agents\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Fgenerated\u002Fgemini\u002Fskills\u002F\u003Cskill>\u002FSKILL.md` | Shared native skill and generated copy |\n| **Claude Code** | `.claude\u002Fskills\u002F\u003Cskill>\u002FSKILL.md` | Native on-demand skill |\n| **GitHub Copilot** | `.github\u002Fcopilot-instructions.md` | Managed block with source paths and routing descriptions |\n| **Cursor** | `.cursor\u002Frules\u002F\u003Cskill>.mdc` | MDC rule with description, globs, and alwaysApply |\n| **Aider** | `.aider.conf.yml`, `CONVENTIONS.md` | YAML read entry for conventions; on-demand source references |\n| **Zed IDE** | `.zed\u002Frules.md`, `.zed\u002Fprompts\u002F\u003Cskill>.md` | Rule index and separate prompt body |\n\nWhen `.agents\u002FAGENTS.md` exists, Cursor also emits\n`.cursor\u002Frules\u002F00-project-rules.mdc`, Zed emits `.zed\u002Fprompts\u002Fcontextos.md`,\nand Copilot, Aider, and Zed include project rules in their instruction files.\nClaude does not generate `CLAUDE.md` or `.agents\u002Fgenerated\u002Fclaude\u002F`;\nCursor does not generate a legacy `.cursorrules` fallback.\n\n## Activation\n\nSet `activation` and `globs` in the winning manifest or entrypoint frontmatter.\nAbsent both, known catalog path scopes apply; otherwise the default is `auto`.\nThe resolver's `signals.fileGlobs` affects recommendations, not adapter activation.\n\n| Target | Activation | Native artifacts | Metadata \u002F behavior |\n| --- | --- | --- | --- |\n| `gemini` | `auto` | `.agents\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Fgenerated\u002Fgemini\u002Fskills\u002F\u003Cskill>\u002FSKILL.md` | Description-based discovery; no activation rule |\n| `gemini` | `always` | `.agents\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Fgenerated\u002Fgemini\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Frules\u002Fcontextos-\u003Cskill>.md` | Rule trigger is always_on |\n| `gemini` | `globs` | `.agents\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Fgenerated\u002Fgemini\u002Fskills\u002F\u003Cskill>\u002FSKILL.md`, `.agents\u002Frules\u002Fcontextos-\u003Cskill>.md` | Rule trigger is glob; additional bounded Windows rule files may be emitted |\n| `claude` | `auto` | `.claude\u002Fskills\u002F\u003Cskill>\u002FSKILL.md` | Name and description frontmatter; body on demand |\n| `claude` | `always` | `.claude\u002Frules\u002F\u003Cskill>.md` | Full body without paths frontmatter |\n| `claude` | `globs` | `.claude\u002Frules\u002F\u003Cskill>.md` | Full body with paths list |\n| `cursor` | `auto` | `.cursor\u002Frules\u002F\u003Cskill>.mdc` | Agent-requested description, empty globs, alwaysApply false |\n| `cursor` | `always` | `.cursor\u002Frules\u002F\u003Cskill>.mdc` | Empty globs, alwaysApply true |\n| `cursor` | `globs` | `.cursor\u002Frules\u002F\u003Cskill>.mdc` | Declared globs, alwaysApply false |\n\nCopilot, Aider, and Zed inline `always` skill bodies in their instruction files.\nOther skills remain routing references there; Zed also emits every skill's prompt\nbody. These three adapters do not enforce per-file glob activation. Markdown\nresource links are relocated when entrypoints move; load supporting files only\nwhen needed. Export selects installed skills allowed by the profile, not just the\nresolver's recommendation for the current task.\n\n## Generation Commands\n\n- `contextos compile` validates manifests and rebuilds the registry.\n- `contextos export \u003Cagent>` renders and applies one of the six targets above.\n- `contextos export all` applies all adapters in one transaction.\n- `contextos export all --check --json` checks drift without writing outputs.\n- `contextos export all --dry-run --json` inspects planned artifacts.\n\nIn this source checkout, use `node .agents\u002Fctx.js` instead of `contextos`.\nAfter source edits, compile, export, then check. A successful check means the\nselected projections match their sources and recorded ownership; it does not\nprove that a live client loaded them. Review ownership conflicts instead of\nautomatically forcing replacement. Aider merges its YAML read list and preserves\nexisting settings; inspect new defaults before running the client.\n\n## Code Examples\n\nSee [EXAMPLES.md](EXAMPLES.md) for the default adapters skill outputs and a\npath-scoped activation example.\n\n## Validation Checklist\n\n- [ ] Correct project root, winning manifest, entrypoint, resources, and profile.\n- [ ] Expected native path and activation metadata for each selected client.\n- [ ] Compile succeeds, export preserves user changes, and check reports no drift.\n- [ ] Generated-file evidence and any live-client testing are reported separately.\n\n## Common Mistakes\n\nEditing outputs, confusing resolver selection with export scope, assuming every\nadapter implements file globs, or treating a generated file as model-quality proof.\nDo not expect Continue or OpenHands exports; they are not registered targets.\n\n## Integration Notes\n\nUse context-os for manifest, resolver, profile, and budget configuration;\nengineering-workflow owns proportional implementation and verification.\n",{"name":5,"description":7},"main","7fb8666fa73d9b8f9707a47505553f0f1af474ea","catalog\u002Fskills\u002Fadapters",1,"https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents","https:\u002F\u002Fgithub.com\u002Fkok-o","nurkhan.space","Software engineer focused on AI tooling and ML infrastructure. \r\n\r\n","2026-11-11 02:56:50",{"source":8,"skillName":5,"commands":34,"prompt":44,"zipAvailable":45},[35,38,41],{"tool":36,"command":37},"npx","npx skills add kok-o\u002Fcontextos-agents --skill adapters",{"tool":39,"command":40},"bunx","bunx skills add kok-o\u002Fcontextos-agents --skill adapters",{"tool":42,"command":43},"pnpm","pnpm dlx skills add kok-o\u002Fcontextos-agents --skill adapters","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Fcatalog\u002Fskills\u002Fadapters\u002FSKILL.md",true,[47,53,57,61],{"path":48,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":51,"sourceHtmlUrl":52},"EXAMPLES.md","REFERENCE","参考文档",2107,"https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Fcatalog\u002Fskills\u002Fadapters\u002FEXAMPLES.md",{"path":54,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":55,"sourceHtmlUrl":56},"skill.yaml",342,"https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Fcatalog\u002Fskills\u002Fadapters\u002Fskill.yaml",{"path":58,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":59,"sourceHtmlUrl":60},"TROUBLESHOOTING.md",308,"https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Fcatalog\u002Fskills\u002Fadapters\u002FTROUBLESHOOTING.md",{"path":62,"resourceType":49,"resourceTypeDesc":50,"sizeBytes":63,"sourceHtmlUrl":64},"VALIDATION.json",198,"https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Fcatalog\u002Fskills\u002Fadapters\u002FVALIDATION.json",[66,73,80,87,94,100],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":70,"categories":71,"keywords":72,"publishedAt":19},"01M78F25RVZVK4KYXZZKW42DZH","scoped-frontend","Frontend component design and React guidelines","https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fadapter-contracts\u002F.agents\u002Fcore\u002Fskills\u002Fscoped-frontend\u002FSKILL.md",[],[],{"skillNo":74,"name":75,"nameZh":6,"description":76,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":77,"categories":78,"keywords":79,"publishedAt":19},"01M78F23169Y0R4CB819166JZP","scoped-backend","Backend service architecture and API endpoints","https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fadapter-contracts\u002F.agents\u002Fcore\u002Fskills\u002Fscoped-backend\u002FSKILL.md",[],[],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":19},"01M78F1ZSMZAMY7S2ECNB065QR","engineering-workflow","Core software engineering lifecycle workflow","https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Ftests\u002Ffixtures\u002Fadapter-contracts\u002F.agents\u002Fcore\u002Fskills\u002Fengineering-workflow\u002FSKILL.md",[],[],{"skillNo":88,"name":89,"nameZh":6,"description":90,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":91,"categories":92,"keywords":93,"publishedAt":19},"01M78F1WWD36GB27TAS4TSM3MJ","team-auth","ContextOS skill for team-auth","https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Fscripts\u002Freadme-gif\u002Fevidence\u002Fexports\u002Fcodex\u002FSKILL.md",[],[],{"skillNo":95,"name":96,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"publishedAt":19},"01M78F1TEZPGE4NCTPZRC7Q1KK","claude","https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Fscripts\u002Freadme-gif\u002Fevidence\u002Fexports\u002Fclaude\u002FSKILL.md",[],[],{"skillNo":101,"name":102,"nameZh":6,"description":103,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":104,"categories":105,"keywords":106,"publishedAt":19},"01M78F1Q8JG229RWD6ESVKT9R9","team-order","Team conventions for createOrder and item name normalization in src\u002Forder.ts.","https:\u002F\u002Fgithub.com\u002Fkok-o\u002Fcontextos-agents\u002Fblob\u002Fmain\u002Fexamples\u002Freadme-demo\u002Fteam-order\u002FSKILL.md",[],[]]