[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M79K4R6GPK8XGN6NHYA3RSTP":3,"skill-same:01M79K4R6GPK8XGN6NHYA3RSTP":44},{"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":14,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M79K4R6GPK8XGN6NHYA3RSTP","record-and-replay",null,"Use when the user wants Codex to record a Linux desktop or browser workflow and turn it into a reusable skill. Requires the Record & Replay event-stream MCP server.","aleafofwutong\u002Fcodex_for_ubuntu","aleafofwutong","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F183128716?v=4","01M79K4D17B01YHEWF0XJQ03PB","firmiana","01M79K4D1R6SNV4KE4MZFH0S53",0,"MIT","https:\u002F\u002Fgithub.com\u002Faleafofwutong\u002Fcodex_for_ubuntu\u002Fblob\u002Fmain\u002Flinux-features\u002Frecord-and-replay\u002Fplugin-template\u002Fskills\u002Frecord-and-replay\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: record-and-replay\ndescription: \"Use when the user wants Codex to record a Linux desktop or browser workflow and turn it into a reusable skill. Requires the Record & Replay event-stream MCP server.\"\n---\n\n# Record & Replay\n\nRecord the user's Linux workflow, inspect the event stream, and turn the\nrepeatable intent into a reusable Codex skill. This is the same bundled\nRecord & Replay product shell backed by a Linux-native event-stream server,\nnot a raw pointer or keyboard macro recorder.\n\nThe bundled plugin launches `.\u002Fbin\u002FSkyLinuxComputerUseClient event-stream mcp`.\nThat helper is the Linux equivalent of the macOS Sky Computer Use event-stream\nclient, implemented by the Rust `codex-record-replay-linux` backend.\n\n## Flow\n\n1. Call `doctor` when readiness is uncertain. If Computer Use cannot observe the\n   desktop, explain the blocker and offer the diagnostics\u002Fsetup path before\n   recording.\n2. Call `skysight_start` when recent activity context should keep accumulating\n   before or during the demo, or `skysight_snapshot` when you only need a\n   point-in-time local activity summary. Use `skysight_pause` and\n   `skysight_resume` to stop or continue Chronicle-compatible resources\n   without losing the active session, and use `skysight_status` to find the\n   resource paths and local OCR availability. Respect\n   `skysight_list_exclusions` and update exclusions before recording sensitive\n   apps or domains. Treat OCR as local screen evidence metadata; do not copy raw\n   OCR text into durable skill drafts unless it is necessary for the reusable\n   workflow and safe to retain.\n3. Call `event_stream_start` with a short `goal` when matching the upstream\n   Record & Replay flow, or `start` when you need Linux-specific options. The\n   Linux app should show the active Record & Replay recording HUD while the\n   shared runtime status is active. Native Linux audio evidence is opt-in and\n   requires both `include_audio: true` and an affirmative\n   `CODEX_RECORD_REPLAY_AUDIO` setting; normal workflow recording should rely\n   on transcript `speech_context` instead.\n   Tell the user that recording is active, that they should perform the workflow\n   normally, and that they can say `done` when finished.\n4. During the demonstration, call `desktop_snapshot` at meaningful app\u002Fwindow\n   changes, such as after the user opens Chrome or arrives on a target site.\n   Call `mark` only for meaningful intent boundaries that will help the future\n   skill, such as \"source page opened\" or \"finished selecting rows\".\n5. When transcript text is explicitly available during the recording, call\n   `speech_context` with the transcript. Treat the speech as user\n   intent\u002Fcontext, not as audio to replay or Chronicle-compatible resources.\n   Do not hijack the composer dictation UI as the recording architecture.\n6. For browser workflows, call `browser_trace` when browser\u002FCDP trace evidence\n   is available. Treat the trace as semantic evidence for drafting the skill,\n   not as a click\u002Fcoordinate replay script.\n7. Use `event_stream_status` or `status` if you need to confirm which bundle is\n   active. When the user says they are done, asks to stop, or the HUD sends\n   \"I'm done recording.\", call `event_stream_stop` or `stop` if the bundle is\n   still active.\n8. If the user discards the recording or the HUD cancel control is used, call\n   `event_stream_cancel` or `cancel` with `discarded: true` and treat the bundle\n   as canceled evidence only.\n9. Call `validate_bundle`, then `draft_skill_prompt`.\n10. Use the draft prompt and the bundle evidence to create or update a normal\n   `SKILL.md`. Prefer stable app names, URLs, semantic UI labels, and data\n   shape descriptions over literal coordinates.\n11. Call `inspect_skill` before import. Call `import_skill` only after the user\n   approves the generated skill.\n\n## Guardrails\n\n- Do not replay raw mouse coordinates or keystroke timing as the primary plan.\n- Avoid exposing private captured content unless it is needed to describe the\n  reusable workflow shape. This includes spoken transcript context.\n- Keep generated skills source-aware: prefer browser DOM, APIs, files, and\n  named UI controls before visual matching.\n- If the workflow depends on unsupported Linux desktop capabilities, mark the\n  skill conditional and include the Computer Use readiness path.\n","# Record & Replay\n\nRecord the user's Linux workflow, inspect the event stream, and turn the\nrepeatable intent into a reusable Codex skill. This is the same bundled\nRecord & Replay product shell backed by a Linux-native event-stream server,\nnot a raw pointer or keyboard macro recorder.\n\nThe bundled plugin launches `.\u002Fbin\u002FSkyLinuxComputerUseClient event-stream mcp`.\nThat helper is the Linux equivalent of the macOS Sky Computer Use event-stream\nclient, implemented by the Rust `codex-record-replay-linux` backend.\n\n## Flow\n\n1. Call `doctor` when readiness is uncertain. If Computer Use cannot observe the\n   desktop, explain the blocker and offer the diagnostics\u002Fsetup path before\n   recording.\n2. Call `skysight_start` when recent activity context should keep accumulating\n   before or during the demo, or `skysight_snapshot` when you only need a\n   point-in-time local activity summary. Use `skysight_pause` and\n   `skysight_resume` to stop or continue Chronicle-compatible resources\n   without losing the active session, and use `skysight_status` to find the\n   resource paths and local OCR availability. Respect\n   `skysight_list_exclusions` and update exclusions before recording sensitive\n   apps or domains. Treat OCR as local screen evidence metadata; do not copy raw\n   OCR text into durable skill drafts unless it is necessary for the reusable\n   workflow and safe to retain.\n3. Call `event_stream_start` with a short `goal` when matching the upstream\n   Record & Replay flow, or `start` when you need Linux-specific options. The\n   Linux app should show the active Record & Replay recording HUD while the\n   shared runtime status is active. Native Linux audio evidence is opt-in and\n   requires both `include_audio: true` and an affirmative\n   `CODEX_RECORD_REPLAY_AUDIO` setting; normal workflow recording should rely\n   on transcript `speech_context` instead.\n   Tell the user that recording is active, that they should perform the workflow\n   normally, and that they can say `done` when finished.\n4. During the demonstration, call `desktop_snapshot` at meaningful app\u002Fwindow\n   changes, such as after the user opens Chrome or arrives on a target site.\n   Call `mark` only for meaningful intent boundaries that will help the future\n   skill, such as \"source page opened\" or \"finished selecting rows\".\n5. When transcript text is explicitly available during the recording, call\n   `speech_context` with the transcript. Treat the speech as user\n   intent\u002Fcontext, not as audio to replay or Chronicle-compatible resources.\n   Do not hijack the composer dictation UI as the recording architecture.\n6. For browser workflows, call `browser_trace` when browser\u002FCDP trace evidence\n   is available. Treat the trace as semantic evidence for drafting the skill,\n   not as a click\u002Fcoordinate replay script.\n7. Use `event_stream_status` or `status` if you need to confirm which bundle is\n   active. When the user says they are done, asks to stop, or the HUD sends\n   \"I'm done recording.\", call `event_stream_stop` or `stop` if the bundle is\n   still active.\n8. If the user discards the recording or the HUD cancel control is used, call\n   `event_stream_cancel` or `cancel` with `discarded: true` and treat the bundle\n   as canceled evidence only.\n9. Call `validate_bundle`, then `draft_skill_prompt`.\n10. Use the draft prompt and the bundle evidence to create or update a normal\n   `SKILL.md`. Prefer stable app names, URLs, semantic UI labels, and data\n   shape descriptions over literal coordinates.\n11. Call `inspect_skill` before import. Call `import_skill` only after the user\n   approves the generated skill.\n\n## Guardrails\n\n- Do not replay raw mouse coordinates or keystroke timing as the primary plan.\n- Avoid exposing private captured content unless it is needed to describe the\n  reusable workflow shape. This includes spoken transcript context.\n- Keep generated skills source-aware: prefer browser DOM, APIs, files, and\n  named UI controls before visual matching.\n- If the workflow depends on unsupported Linux desktop capabilities, mark the\n  skill conditional and include the Computer Use readiness path.\n",{"name":5,"description":7},"main","87e1b89b85a6a93970e83570d38654cd494c6500","linux-features\u002Frecord-and-replay\u002Fplugin-template\u002Fskills\u002Frecord-and-replay","https:\u002F\u002Fgithub.com\u002Faleafofwutong\u002Fcodex_for_ubuntu","https:\u002F\u002Fgithub.com\u002Faleafofwutong","2026-07-24 08:23:57",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add aleafofwutong\u002Fcodex_for_ubuntu --skill record-and-replay",{"tool":36,"command":37},"bunx","bunx skills add aleafofwutong\u002Fcodex_for_ubuntu --skill record-and-replay",{"tool":39,"command":40},"pnpm","pnpm dlx skills add aleafofwutong\u002Fcodex_for_ubuntu --skill record-and-replay","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Faleafofwutong\u002Fcodex_for_ubuntu\u002Fblob\u002Fmain\u002Flinux-features\u002Frecord-and-replay\u002Fplugin-template\u002Fskills\u002Frecord-and-replay\u002FSKILL.md",true,[],[45,52,59,66,73,80],{"skillNo":46,"name":47,"nameZh":6,"description":48,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":49,"categories":50,"keywords":51,"publishedAt":19},"01M79K50WT51J719YY6PKQNNXX","Unsafe Symlink","Use when testing symlink safety.","https:\u002F\u002Fgithub.com\u002Faleafofwutong\u002Fcodex_for_ubuntu\u002Fblob\u002Fmain\u002Frecord-replay-linux\u002Ftests\u002Ffixtures\u002Fskills\u002Funsafe-symlink\u002FSKILL.md",[],[],{"skillNo":53,"name":54,"nameZh":6,"description":55,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":56,"categories":57,"keywords":58,"publishedAt":19},"01M79K4ZW2XHAXS7DZYGYA1Z17","Mac Only","Use when automating Finder with AppleScript.","https:\u002F\u002Fgithub.com\u002Faleafofwutong\u002Fcodex_for_ubuntu\u002Fblob\u002Fmain\u002Frecord-replay-linux\u002Ftests\u002Ffixtures\u002Fskills\u002Fplatform-macos\u002FSKILL.md",[],[],{"skillNo":60,"name":61,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":19},"01M79K4XNVSNB063QDKVVN0Y45","Instruction Only","Use when answering with reusable written instructions.","https:\u002F\u002Fgithub.com\u002Faleafofwutong\u002Fcodex_for_ubuntu\u002Fblob\u002Fmain\u002Frecord-replay-linux\u002Ftests\u002Ffixtures\u002Fskills\u002Finstruction-only\u002FSKILL.md",[],[],{"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},"01M79K4W2TE0KB7QJ7Z01VA6H5","Desktop Act","Use when clicking and typing in a Linux desktop app.","https:\u002F\u002Fgithub.com\u002Faleafofwutong\u002Fcodex_for_ubuntu\u002Fblob\u002Fmain\u002Frecord-replay-linux\u002Ftests\u002Ffixtures\u002Fskills\u002Fdesktop-act\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},"01M79K4V397B122QP3YKQ1RHWZ","Collision Tripwire","Use when testing script handling.","https:\u002F\u002Fgithub.com\u002Faleafofwutong\u002Fcodex_for_ubuntu\u002Fblob\u002Fmain\u002Frecord-replay-linux\u002Ftests\u002Ffixtures\u002Fskills\u002Fcollision-tripwire\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},"01M79K4KV8JS8DXMZC5YCF67XJ","agent-workspace-linux","Use when a task needs an isolated hidden Linux desktop or workspace-owned browser: GUI app QA, web\u002Fbrowser\u002Fshopping automation, sandboxed app observation, or stale workspace cleanup. Routes agent-workspace-linux MCP tools on demand. Does NOT apply to host desktop\u002FChrome control, generic MCP setup, or pure code\u002Ffile edits.","https:\u002F\u002Fgithub.com\u002Faleafofwutong\u002Fcodex_for_ubuntu\u002Fblob\u002Fmain\u002Flinux-features\u002Fagent-workspace\u002Fskills\u002Fagent-workspace-linux\u002FSKILL.md",[],[]]