[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M788W07DKNERHZADFGPBS2FP":3,"skill-same:01M788W07DKNERHZADFGPBS2FP":78},{"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":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M788W07DKNERHZADFGPBS2FP","cli-design",null,"Design and implement trustworthy modern command-line application experiences. Use for CLI UX, developer tools, command output design, progress feedback, status visibility, error messages, interactive CLI flows, AI agent CLI execution views, terminal color tokens, table\u002Flist output, stdout\u002Fstderr behavior, verbose\u002Fdebug output, and machine-readable JSON or structured output modes.","flash-dev-ctrl\u002Fsentra-cli","flash-dev-ctrl","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F300809526?v=4","01M788T0S4NR70G53A5AERVDE3","01M788T0SW4BRHTP0MXK8NVT6P",6,"MIT","https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fcli-design\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: cli-design\ndescription: Design and implement trustworthy modern command-line application experiences. Use for CLI UX, developer tools, command output design, progress feedback, status visibility, error messages, interactive CLI flows, AI agent CLI execution views, terminal color tokens, table\u002Flist output, stdout\u002Fstderr behavior, verbose\u002Fdebug output, and machine-readable JSON or structured output modes.\n---\n\n# CLI Design\n\nUse this skill to design CLI applications that feel visible, understandable, and trustworthy. Treat output as an interaction system, not decoration: every line should help the user know what is happening, what changed, whether the tool is alive, and what they can do next.\n\n## Design Workflow\n\n1. Identify the command type: one-shot, long-running, interactive, automation-first, or AI-agent workflow.\n2. Model trust questions: Is it alive? What is it doing? Why is it doing that? How far along is it? How do I recover or continue?\n3. Design information architecture: Primary, Secondary, Metadata.\n4. Compose human output from line roles before writing text.\n5. Select output modes: default, verbose, machine.\n6. Apply semantic feedback: status, progress, result, next step.\n7. Validate against the anti-pattern checklist before implementing.\n\nThe default interaction structure is:\n\n```text\nContext\nAction\nProgress\nResult\nNext Step\n```\n\n## Reference Routing\n\nLoad only the reference files needed for the task:\n\n| Task | Read |\n| --- | --- |\n| Overall CLI UX strategy, HCI, Unix principles, IA, disclosure | `references\u002Fdesign-principles.md` |\n| Human output templates, stdout\u002Fstderr, default\u002Fverbose\u002Fmachine modes | `references\u002Foutput-patterns.md` |\n| Terminal colors, design tokens, dark theme, `NO_COLOR`, semantic color use | `references\u002Fcolor-system.md` |\n| Status symbols, spinners, long-running tasks, phase progress | `references\u002Fstatus-progress.md` |\n| Error messages, confirmations, keyboard interaction, dangerous actions | `references\u002Ferrors-interaction.md` |\n| Tables, lists, alignment, truncation, JSON alternatives | `references\u002Ftables-lists.md` |\n| AI agent CLI state, planning\u002Ftool\u002Fverification views, approval and sandbox feedback | `references\u002Fai-agent-cli.md` |\n\n## Core Rules\n\n- Compose output from line roles before writing text: headline, metadata, phase, result, next step, and diagnostic.\n- Group context, progress, result, and next step with stable spacing. Use two-space indentation for supporting human-output lines.\n- Never let long-running work stay silent. Show a spinner, phase transition, heartbeat, or progress indicator within a short human-noticeable delay.\n- Design against idle aversion: when work takes time, continuously prove the program is alive with useful status updates, streaming output, or a spinner.\n- Make operations transparent: show phase, action, goal, and necessary context instead of waiting silently and revealing only the final result.\n- Use goal-gradient feedback: as completion approaches, make remaining work and completed phases increasingly visible.\n- Prefer phase-based progress over noisy percentage updates unless the percentage is accurate and useful.\n- Use semantic color only. Color communicates meaning: success, warning, error, info, metadata, focus, or selection.\n- Design for visual comfort in long terminal sessions. Avoid pure-white foreground, all-bold output, high-saturation status colors, and high-contrast glare.\n- Never rely on color alone. Pair color with symbols, labels, layout, or typography.\n- Separate human output from machine output. Machine mode must use a stable structured schema and avoid ANSI styling.\n- Preserve scriptability. Keep intended data on stdout; put progress, warnings, diagnostics, and prompts on stderr.\n- Make errors actionable with `Problem -> Cause -> Solution`.\n- Design dangerous operations with explicit confirmation and automation-safe flags.\n- Provide Unicode status symbols by default, but require ASCII fallback for CI, logs, non-UTF-8 terminals, non-TTY output, and machine mode.\n\n## Anti-Pattern Checklist\n\nReject designs that:\n\n- Print only `Processing...` for a long operation.\n- Produce flat-stack output where every line has equal visual weight.\n- Mix unindented metadata, progress, and result lines in one block.\n- Use gerund-style counted phases such as `[1\u002F2] trying npm`; use imperative actions such as `[1\u002F2] Try npm install`.\n- Print duplicate completion messages such as a success headline followed by `Installed opencode`.\n- Leave the screen static during model inference, network calls, scans, builds, or other opaque work.\n- Perform opaque work with no phase, goal, or current action.\n- Hide operational facts such as `[3\u002F12] scanning ~\u002F.codex\u002Fskills, found 7 skills` until the command ends.\n- Let a progress bar stall in the final 20% without phase, heartbeat, or explanation.\n- Use decorative color or color-only meaning.\n- Render most output as bright white or bold highlighted text.\n- Use color palettes that feel neon, harsh, or visually tiring on dark terminals.\n- Apply semantic color to entire sentences, paragraphs, or blocks instead of symbols, labels, or critical words.\n- Let warning yellow dominate the screen.\n- Use ANSI bright variants as the default foreground or default status palette.\n- Mix progress text into stdout where scripts expect parseable results.\n- Dump raw logs by default instead of summarizing what matters.\n- Hide recovery instructions behind generic errors like `Error: failed`.\n- Use unstable JSON fields or human prose in machine mode.\n- Ask for confirmation in automation mode without an explicit non-interactive alternative.\n- Treat full-screen TUI concerns as ordinary CLI output. For Ratatui, Ink, Textual, Bubbletea, or persistent terminal applications, combine this skill with a TUI design skill.\n","# CLI Design\n\nUse this skill to design CLI applications that feel visible, understandable, and trustworthy. Treat output as an interaction system, not decoration: every line should help the user know what is happening, what changed, whether the tool is alive, and what they can do next.\n\n## Design Workflow\n\n1. Identify the command type: one-shot, long-running, interactive, automation-first, or AI-agent workflow.\n2. Model trust questions: Is it alive? What is it doing? Why is it doing that? How far along is it? How do I recover or continue?\n3. Design information architecture: Primary, Secondary, Metadata.\n4. Compose human output from line roles before writing text.\n5. Select output modes: default, verbose, machine.\n6. Apply semantic feedback: status, progress, result, next step.\n7. Validate against the anti-pattern checklist before implementing.\n\nThe default interaction structure is:\n\n```text\nContext\nAction\nProgress\nResult\nNext Step\n```\n\n## Reference Routing\n\nLoad only the reference files needed for the task:\n\n| Task | Read |\n| --- | --- |\n| Overall CLI UX strategy, HCI, Unix principles, IA, disclosure | `references\u002Fdesign-principles.md` |\n| Human output templates, stdout\u002Fstderr, default\u002Fverbose\u002Fmachine modes | `references\u002Foutput-patterns.md` |\n| Terminal colors, design tokens, dark theme, `NO_COLOR`, semantic color use | `references\u002Fcolor-system.md` |\n| Status symbols, spinners, long-running tasks, phase progress | `references\u002Fstatus-progress.md` |\n| Error messages, confirmations, keyboard interaction, dangerous actions | `references\u002Ferrors-interaction.md` |\n| Tables, lists, alignment, truncation, JSON alternatives | `references\u002Ftables-lists.md` |\n| AI agent CLI state, planning\u002Ftool\u002Fverification views, approval and sandbox feedback | `references\u002Fai-agent-cli.md` |\n\n## Core Rules\n\n- Compose output from line roles before writing text: headline, metadata, phase, result, next step, and diagnostic.\n- Group context, progress, result, and next step with stable spacing. Use two-space indentation for supporting human-output lines.\n- Never let long-running work stay silent. Show a spinner, phase transition, heartbeat, or progress indicator within a short human-noticeable delay.\n- Design against idle aversion: when work takes time, continuously prove the program is alive with useful status updates, streaming output, or a spinner.\n- Make operations transparent: show phase, action, goal, and necessary context instead of waiting silently and revealing only the final result.\n- Use goal-gradient feedback: as completion approaches, make remaining work and completed phases increasingly visible.\n- Prefer phase-based progress over noisy percentage updates unless the percentage is accurate and useful.\n- Use semantic color only. Color communicates meaning: success, warning, error, info, metadata, focus, or selection.\n- Design for visual comfort in long terminal sessions. Avoid pure-white foreground, all-bold output, high-saturation status colors, and high-contrast glare.\n- Never rely on color alone. Pair color with symbols, labels, layout, or typography.\n- Separate human output from machine output. Machine mode must use a stable structured schema and avoid ANSI styling.\n- Preserve scriptability. Keep intended data on stdout; put progress, warnings, diagnostics, and prompts on stderr.\n- Make errors actionable with `Problem -> Cause -> Solution`.\n- Design dangerous operations with explicit confirmation and automation-safe flags.\n- Provide Unicode status symbols by default, but require ASCII fallback for CI, logs, non-UTF-8 terminals, non-TTY output, and machine mode.\n\n## Anti-Pattern Checklist\n\nReject designs that:\n\n- Print only `Processing...` for a long operation.\n- Produce flat-stack output where every line has equal visual weight.\n- Mix unindented metadata, progress, and result lines in one block.\n- Use gerund-style counted phases such as `[1\u002F2] trying npm`; use imperative actions such as `[1\u002F2] Try npm install`.\n- Print duplicate completion messages such as a success headline followed by `Installed opencode`.\n- Leave the screen static during model inference, network calls, scans, builds, or other opaque work.\n- Perform opaque work with no phase, goal, or current action.\n- Hide operational facts such as `[3\u002F12] scanning ~\u002F.codex\u002Fskills, found 7 skills` until the command ends.\n- Let a progress bar stall in the final 20% without phase, heartbeat, or explanation.\n- Use decorative color or color-only meaning.\n- Render most output as bright white or bold highlighted text.\n- Use color palettes that feel neon, harsh, or visually tiring on dark terminals.\n- Apply semantic color to entire sentences, paragraphs, or blocks instead of symbols, labels, or critical words.\n- Let warning yellow dominate the screen.\n- Use ANSI bright variants as the default foreground or default status palette.\n- Mix progress text into stdout where scripts expect parseable results.\n- Dump raw logs by default instead of summarizing what matters.\n- Hide recovery instructions behind generic errors like `Error: failed`.\n- Use unstable JSON fields or human prose in machine mode.\n- Ask for confirmation in automation mode without an explicit non-interactive alternative.\n- Treat full-screen TUI concerns as ordinary CLI output. For Ratatui, Ink, Textual, Bubbletea, or persistent terminal applications, combine this skill with a TUI design skill.\n",{"name":5,"description":7},"main","122ca2a56b19fb97376c050637b7c5ce8aaa3793",".codex\u002Fskills\u002Fcli-design",0,"https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli","https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl","2026-08-17 19:17:34",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add flash-dev-ctrl\u002Fsentra-cli --skill cli-design",{"tool":36,"command":37},"bunx","bunx skills add flash-dev-ctrl\u002Fsentra-cli --skill cli-design",{"tool":39,"command":40},"pnpm","pnpm dlx skills add flash-dev-ctrl\u002Fsentra-cli --skill cli-design","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fcli-design\u002FSKILL.md",true,[44,50,54,58,62,66,70,74],{"path":45,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":48,"sourceHtmlUrl":49},"agents\u002Fopenai.yaml","REFERENCE","参考文档",223,"https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fcli-design\u002Fagents\u002Fopenai.yaml",{"path":51,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":52,"sourceHtmlUrl":53},"references\u002Fai-agent-cli.md",2904,"https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fcli-design\u002Freferences\u002Fai-agent-cli.md",{"path":55,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":56,"sourceHtmlUrl":57},"references\u002Fcolor-system.md",14140,"https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fcli-design\u002Freferences\u002Fcolor-system.md",{"path":59,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":60,"sourceHtmlUrl":61},"references\u002Fdesign-principles.md",3608,"https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fcli-design\u002Freferences\u002Fdesign-principles.md",{"path":63,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":64,"sourceHtmlUrl":65},"references\u002Ferrors-interaction.md",2477,"https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fcli-design\u002Freferences\u002Ferrors-interaction.md",{"path":67,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":68,"sourceHtmlUrl":69},"references\u002Foutput-patterns.md",7076,"https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fcli-design\u002Freferences\u002Foutput-patterns.md",{"path":71,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":72,"sourceHtmlUrl":73},"references\u002Fstatus-progress.md",3820,"https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fcli-design\u002Freferences\u002Fstatus-progress.md",{"path":75,"resourceType":46,"resourceTypeDesc":47,"sizeBytes":76,"sourceHtmlUrl":77},"references\u002Ftables-lists.md",3948,"https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002F.codex\u002Fskills\u002Fcli-design\u002Freferences\u002Ftables-lists.md",[79,86,93,100,107],{"skillNo":80,"name":81,"nameZh":6,"description":82,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":83,"categories":84,"keywords":85,"publishedAt":18},"01M788YD3R630P5T6F8AVF3RF3","test-ti","A test skill for security scanning","https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002Ffixtures\u002Fskill\u002Ftest-ti\u002FSKILL.md",[],[],{"skillNo":87,"name":88,"nameZh":6,"description":89,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":90,"categories":91,"keywords":92,"publishedAt":18},"01M788Y3556ASFKR43W4RPPHG3","rogue-deploy","Deploys application to remote servers","https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002Ffixtures\u002Fskill\u002Frogue-deploy\u002FSKILL.md",[],[],{"skillNo":94,"name":95,"nameZh":6,"description":96,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":97,"categories":98,"keywords":99,"publishedAt":18},"01M788XMCPEJSTMW56A32EGTZ9","prompt-hijack","A helper skill that boosts productivity","https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002Ffixtures\u002Fskill\u002Fprompt-hijack\u002FSKILL.md",[],[],{"skillNo":101,"name":102,"nameZh":6,"description":103,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":104,"categories":105,"keywords":106,"publishedAt":18},"01M788WYY1WD4KBVMGJEW5MS5W","math-calculator","This skill should be used when Claude Code needs to perform basic arithmetic calculations. It provides a Python script that safely evaluates mathematical expressions including addition, subtraction, multiplication, division, exponentiation, and square roots.","https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002Ffixtures\u002Fskill\u002Fmath-calculator\u002FSKILL.md",[],[],{"skillNo":108,"name":109,"nameZh":6,"description":110,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":111,"categories":112,"keywords":113,"publishedAt":18},"01M788WEERCBBW223QT56A27VH","hack-skill","A suspicious skill","https:\u002F\u002Fgithub.com\u002Fflash-dev-ctrl\u002Fsentra-cli\u002Fblob\u002Fmain\u002Ffixtures\u002Fskill\u002Fhack\u002FSKILL.md",[],[]]