[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-vercel-labs-skill-creator":3,"mdc--yiu3cr-key":33,"related-repo-vercel-labs-skill-creator":2023,"related-org-vercel-labs-skill-creator":2127},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":22,"repoUrl":23,"updatedAt":24,"license":25,"forks":26,"topics":27,"repo":28,"sourceUrl":31,"mdContent":32},"skill-creator","author and update agent skills","Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"vercel-labs","Vercel Labs","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fvercel-labs.png",[12,16,19],{"name":13,"slug":14,"type":15},"Documentation","documentation","tag",{"name":17,"slug":18,"type":15},"Plugin Development","plugin-development",{"name":20,"slug":21,"type":15},"Technical Writing","technical-writing",15678,"https:\u002F\u002Fgithub.com\u002Fvercel-labs\u002Fjson-render","2026-07-17T06:08:37.425038","Complete terms in LICENSE.txt",845,[],{"repoUrl":23,"stars":22,"forks":26,"topics":29,"description":30},[],"The Generative UI framework","https:\u002F\u002Fgithub.com\u002Fvercel-labs\u002Fjson-render\u002Ftree\u002FHEAD\u002Fskills\u002Fskill-creator","---\nname: skill-creator\ndescription: Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.\nlicense: Complete terms in LICENSE.txt\n---\n\n# Skill Creator\n\nThis skill provides guidance for creating effective skills.\n\n## About Skills\n\nSkills are modular, self-contained packages that extend Claude's capabilities by providing\nspecialized knowledge, workflows, and tools. Think of them as \"onboarding guides\" for specific\ndomains or tasks—they transform Claude from a general-purpose agent into a specialized agent\nequipped with procedural knowledge that no model can fully possess.\n\n### What Skills Provide\n\n1. Specialized workflows - Multi-step procedures for specific domains\n2. Tool integrations - Instructions for working with specific file formats or APIs\n3. Domain expertise - Company-specific knowledge, schemas, business logic\n4. Bundled resources - Scripts, references, and assets for complex and repetitive tasks\n\n## Core Principles\n\n### Concise is Key\n\nThe context window is a public good. Skills share the context window with everything else Claude needs: system prompt, conversation history, other Skills' metadata, and the actual user request.\n\n**Default assumption: Claude is already very smart.** Only add context Claude doesn't already have. Challenge each piece of information: \"Does Claude really need this explanation?\" and \"Does this paragraph justify its token cost?\"\n\nPrefer concise examples over verbose explanations.\n\n### Set Appropriate Degrees of Freedom\n\nMatch the level of specificity to the task's fragility and variability:\n\n**High freedom (text-based instructions)**: Use when multiple approaches are valid, decisions depend on context, or heuristics guide the approach.\n\n**Medium freedom (pseudocode or scripts with parameters)**: Use when a preferred pattern exists, some variation is acceptable, or configuration affects behavior.\n\n**Low freedom (specific scripts, few parameters)**: Use when operations are fragile and error-prone, consistency is critical, or a specific sequence must be followed.\n\nThink of Claude as exploring a path: a narrow bridge with cliffs needs specific guardrails (low freedom), while an open field allows many routes (high freedom).\n\n### Anatomy of a Skill\n\nEvery skill consists of a required SKILL.md file and optional bundled resources:\n\n```\nskill-name\u002F\n├── SKILL.md (required)\n│   ├── YAML frontmatter metadata (required)\n│   │   ├── name: (required)\n│   │   └── description: (required)\n│   └── Markdown instructions (required)\n└── Bundled Resources (optional)\n    ├── scripts\u002F          - Executable code (Python\u002FBash\u002Fetc.)\n    ├── references\u002F       - Documentation intended to be loaded into context as needed\n    └── assets\u002F           - Files used in output (templates, icons, fonts, etc.)\n```\n\n#### SKILL.md (required)\n\nEvery SKILL.md consists of:\n\n- **Frontmatter** (YAML): Contains `name` and `description` fields. These are the only fields that Claude reads to determine when the skill gets used, thus it is very important to be clear and comprehensive in describing what the skill is, and when it should be used.\n- **Body** (Markdown): Instructions and guidance for using the skill. Only loaded AFTER the skill triggers (if at all).\n\n#### Bundled Resources (optional)\n\n##### Scripts (`scripts\u002F`)\n\nExecutable code (Python\u002FBash\u002Fetc.) for tasks that require deterministic reliability or are repeatedly rewritten.\n\n- **When to include**: When the same code is being rewritten repeatedly or deterministic reliability is needed\n- **Example**: `scripts\u002Frotate_pdf.py` for PDF rotation tasks\n- **Benefits**: Token efficient, deterministic, may be executed without loading into context\n- **Note**: Scripts may still need to be read by Claude for patching or environment-specific adjustments\n\n##### References (`references\u002F`)\n\nDocumentation and reference material intended to be loaded as needed into context to inform Claude's process and thinking.\n\n- **When to include**: For documentation that Claude should reference while working\n- **Examples**: `references\u002Ffinance.md` for financial schemas, `references\u002Fmnda.md` for company NDA template, `references\u002Fpolicies.md` for company policies, `references\u002Fapi_docs.md` for API specifications\n- **Use cases**: Database schemas, API documentation, domain knowledge, company policies, detailed workflow guides\n- **Benefits**: Keeps SKILL.md lean, loaded only when Claude determines it's needed\n- **Best practice**: If files are large (>10k words), include grep search patterns in SKILL.md\n- **Avoid duplication**: Information should live in either SKILL.md or references files, not both. Prefer references files for detailed information unless it's truly core to the skill—this keeps SKILL.md lean while making information discoverable without hogging the context window. Keep only essential procedural instructions and workflow guidance in SKILL.md; move detailed reference material, schemas, and examples to references files.\n\n##### Assets (`assets\u002F`)\n\nFiles not intended to be loaded into context, but rather used within the output Claude produces.\n\n- **When to include**: When the skill needs files that will be used in the final output\n- **Examples**: `assets\u002Flogo.png` for brand assets, `assets\u002Fslides.pptx` for PowerPoint templates, `assets\u002Ffrontend-template\u002F` for HTML\u002FReact boilerplate, `assets\u002Ffont.ttf` for typography\n- **Use cases**: Templates, images, icons, boilerplate code, fonts, sample documents that get copied or modified\n- **Benefits**: Separates output resources from documentation, enables Claude to use files without loading them into context\n\n#### What to Not Include in a Skill\n\nA skill should only contain essential files that directly support its functionality. Do NOT create extraneous documentation or auxiliary files, including:\n\n- README.md\n- INSTALLATION_GUIDE.md\n- QUICK_REFERENCE.md\n- CHANGELOG.md\n- etc.\n\nThe skill should only contain the information needed for an AI agent to do the job at hand. It should not contain auxilary context about the process that went into creating it, setup and testing procedures, user-facing documentation, etc. Creating additional documentation files just adds clutter and confusion.\n\n### Progressive Disclosure Design Principle\n\nSkills use a three-level loading system to manage context efficiently:\n\n1. **Metadata (name + description)** - Always in context (~100 words)\n2. **SKILL.md body** - When skill triggers (\u003C5k words)\n3. **Bundled resources** - As needed by Claude (Unlimited because scripts can be executed without reading into context window)\n\n#### Progressive Disclosure Patterns\n\nKeep SKILL.md body to the essentials and under 500 lines to minimize context bloat. Split content into separate files when approaching this limit. When splitting out content into other files, it is very important to reference them from SKILL.md and describe clearly when to read them, to ensure the reader of the skill knows they exist and when to use them.\n\n**Key principle:** When a skill supports multiple variations, frameworks, or options, keep only the core workflow and selection guidance in SKILL.md. Move variant-specific details (patterns, examples, configuration) into separate reference files.\n\n**Pattern 1: High-level guide with references**\n\n```markdown\n# PDF Processing\n\n## Quick start\n\nExtract text with pdfplumber:\n[code example]\n\n## Advanced features\n\n- **Form filling**: See [FORMS.md](FORMS.md) for complete guide\n- **API reference**: See [REFERENCE.md](REFERENCE.md) for all methods\n- **Examples**: See [EXAMPLES.md](EXAMPLES.md) for common patterns\n```\n\nClaude loads FORMS.md, REFERENCE.md, or EXAMPLES.md only when needed.\n\n**Pattern 2: Domain-specific organization**\n\nFor Skills with multiple domains, organize content by domain to avoid loading irrelevant context:\n\n```\nbigquery-skill\u002F\n├── SKILL.md (overview and navigation)\n└── reference\u002F\n    ├── finance.md (revenue, billing metrics)\n    ├── sales.md (opportunities, pipeline)\n    ├── product.md (API usage, features)\n    └── marketing.md (campaigns, attribution)\n```\n\nWhen a user asks about sales metrics, Claude only reads sales.md.\n\nSimilarly, for skills supporting multiple frameworks or variants, organize by variant:\n\n```\ncloud-deploy\u002F\n├── SKILL.md (workflow + provider selection)\n└── references\u002F\n    ├── aws.md (AWS deployment patterns)\n    ├── gcp.md (GCP deployment patterns)\n    └── azure.md (Azure deployment patterns)\n```\n\nWhen the user chooses AWS, Claude only reads aws.md.\n\n**Pattern 3: Conditional details**\n\nShow basic content, link to advanced content:\n\n```markdown\n# DOCX Processing\n\n## Creating documents\n\nUse docx-js for new documents. See [DOCX-JS.md](DOCX-JS.md).\n\n## Editing documents\n\nFor simple edits, modify the XML directly.\n\n**For tracked changes**: See [REDLINING.md](REDLINING.md)\n**For OOXML details**: See [OOXML.md](OOXML.md)\n```\n\nClaude reads REDLINING.md or OOXML.md only when the user needs those features.\n\n**Important guidelines:**\n\n- **Avoid deeply nested references** - Keep references one level deep from SKILL.md. All reference files should link directly from SKILL.md.\n- **Structure longer reference files** - For files longer than 100 lines, include a table of contents at the top so Claude can see the full scope when previewing.\n\n## Skill Creation Process\n\nSkill creation involves these steps:\n\n1. Understand the skill with concrete examples\n2. Plan reusable skill contents (scripts, references, assets)\n3. Initialize the skill (run init_skill.py)\n4. Edit the skill (implement resources and write SKILL.md)\n5. Package the skill (run package_skill.py)\n6. Iterate based on real usage\n\nFollow these steps in order, skipping only if there is a clear reason why they are not applicable.\n\n### Step 1: Understanding the Skill with Concrete Examples\n\nSkip this step only when the skill's usage patterns are already clearly understood. It remains valuable even when working with an existing skill.\n\nTo create an effective skill, clearly understand concrete examples of how the skill will be used. This understanding can come from either direct user examples or generated examples that are validated with user feedback.\n\nFor example, when building an image-editor skill, relevant questions include:\n\n- \"What functionality should the image-editor skill support? Editing, rotating, anything else?\"\n- \"Can you give some examples of how this skill would be used?\"\n- \"I can imagine users asking for things like 'Remove the red-eye from this image' or 'Rotate this image'. Are there other ways you imagine this skill being used?\"\n- \"What would a user say that should trigger this skill?\"\n\nTo avoid overwhelming users, avoid asking too many questions in a single message. Start with the most important questions and follow up as needed for better effectiveness.\n\nConclude this step when there is a clear sense of the functionality the skill should support.\n\n### Step 2: Planning the Reusable Skill Contents\n\nTo turn concrete examples into an effective skill, analyze each example by:\n\n1. Considering how to execute on the example from scratch\n2. Identifying what scripts, references, and assets would be helpful when executing these workflows repeatedly\n\nExample: When building a `pdf-editor` skill to handle queries like \"Help me rotate this PDF,\" the analysis shows:\n\n1. Rotating a PDF requires re-writing the same code each time\n2. A `scripts\u002Frotate_pdf.py` script would be helpful to store in the skill\n\nExample: When designing a `frontend-webapp-builder` skill for queries like \"Build me a todo app\" or \"Build me a dashboard to track my steps,\" the analysis shows:\n\n1. Writing a frontend webapp requires the same boilerplate HTML\u002FReact each time\n2. An `assets\u002Fhello-world\u002F` template containing the boilerplate HTML\u002FReact project files would be helpful to store in the skill\n\nExample: When building a `big-query` skill to handle queries like \"How many users have logged in today?\" the analysis shows:\n\n1. Querying BigQuery requires re-discovering the table schemas and relationships each time\n2. A `references\u002Fschema.md` file documenting the table schemas would be helpful to store in the skill\n\nTo establish the skill's contents, analyze each concrete example to create a list of the reusable resources to include: scripts, references, and assets.\n\n### Step 3: Initializing the Skill\n\nAt this point, it is time to actually create the skill.\n\nSkip this step only if the skill being developed already exists, and iteration or packaging is needed. In this case, continue to the next step.\n\nWhen creating a new skill from scratch, always run the `init_skill.py` script. The script conveniently generates a new template skill directory that automatically includes everything a skill requires, making the skill creation process much more efficient and reliable.\n\nUsage:\n\n```bash\nscripts\u002Finit_skill.py \u003Cskill-name> --path \u003Coutput-directory>\n```\n\nThe script:\n\n- Creates the skill directory at the specified path\n- Generates a SKILL.md template with proper frontmatter and TODO placeholders\n- Creates example resource directories: `scripts\u002F`, `references\u002F`, and `assets\u002F`\n- Adds example files in each directory that can be customized or deleted\n\nAfter initialization, customize or remove the generated SKILL.md and example files as needed.\n\n### Step 4: Edit the Skill\n\nWhen editing the (newly-generated or existing) skill, remember that the skill is being created for another instance of Claude to use. Include information that would be beneficial and non-obvious to Claude. Consider what procedural knowledge, domain-specific details, or reusable assets would help another Claude instance execute these tasks more effectively.\n\n#### Learn Proven Design Patterns\n\nConsult these helpful guides based on your skill's needs:\n\n- **Multi-step processes**: See references\u002Fworkflows.md for sequential workflows and conditional logic\n- **Specific output formats or quality standards**: See references\u002Foutput-patterns.md for template and example patterns\n\nThese files contain established best practices for effective skill design.\n\n#### Start with Reusable Skill Contents\n\nTo begin implementation, start with the reusable resources identified above: `scripts\u002F`, `references\u002F`, and `assets\u002F` files. Note that this step may require user input. For example, when implementing a `brand-guidelines` skill, the user may need to provide brand assets or templates to store in `assets\u002F`, or documentation to store in `references\u002F`.\n\nAdded scripts must be tested by actually running them to ensure there are no bugs and that the output matches what is expected. If there are many similar scripts, only a representative sample needs to be tested to ensure confidence that they all work while balancing time to completion.\n\nAny example files and directories not needed for the skill should be deleted. The initialization script creates example files in `scripts\u002F`, `references\u002F`, and `assets\u002F` to demonstrate structure, but most skills won't need all of them.\n\n#### Update SKILL.md\n\n**Writing Guidelines:** Always use imperative\u002Finfinitive form.\n\n##### Frontmatter\n\nWrite the YAML frontmatter with `name` and `description`:\n\n- `name`: The skill name\n- `description`: This is the primary triggering mechanism for your skill, and helps Claude understand when to use the skill.\n  - Include both what the Skill does and specific triggers\u002Fcontexts for when to use it.\n  - Include all \"when to use\" information here - Not in the body. The body is only loaded after triggering, so \"When to Use This Skill\" sections in the body are not helpful to Claude.\n  - Example description for a `docx` skill: \"Comprehensive document creation, editing, and analysis with support for tracked changes, comments, formatting preservation, and text extraction. Use when Claude needs to work with professional documents (.docx files) for: (1) Creating new documents, (2) Modifying or editing content, (3) Working with tracked changes, (4) Adding comments, or any other document tasks\"\n\nDo not include any other fields in YAML frontmatter.\n\n##### Body\n\nWrite instructions for using the skill and its bundled resources.\n\n### Step 5: Packaging a Skill\n\nOnce development of the skill is complete, it must be packaged into a distributable .skill file that gets shared with the user. The packaging process automatically validates the skill first to ensure it meets all requirements:\n\n```bash\nscripts\u002Fpackage_skill.py \u003Cpath\u002Fto\u002Fskill-folder>\n```\n\nOptional output directory specification:\n\n```bash\nscripts\u002Fpackage_skill.py \u003Cpath\u002Fto\u002Fskill-folder> .\u002Fdist\n```\n\nThe packaging script will:\n\n1. **Validate** the skill automatically, checking:\n\n   - YAML frontmatter format and required fields\n   - Skill naming conventions and directory structure\n   - Description completeness and quality\n   - File organization and resource references\n\n2. **Package** the skill if validation passes, creating a .skill file named after the skill (e.g., `my-skill.skill`) that includes all files and maintains the proper directory structure for distribution. The .skill file is a zip file with a .skill extension.\n\nIf validation fails, the script will report the errors and exit without creating a package. Fix any validation errors and run the packaging command again.\n\n### Step 6: Iterate\n\nAfter testing the skill, users may request improvements. Often this happens right after using the skill, with fresh context of how the skill performed.\n\n**Iteration workflow:**\n\n1. Use the skill on real tasks\n2. Notice struggles or inefficiencies\n3. Identify how SKILL.md or bundled resources should be updated\n4. Implement changes and test again\n",{"data":34,"body":35},{"name":4,"description":6,"license":25},{"type":36,"children":37},"root",[38,46,52,59,64,71,96,102,108,113,124,129,135,140,150,160,170,175,181,186,199,206,211,251,257,272,277,328,341,346,438,451,456,526,532,537,565,570,576,581,614,620,625,635,643,912,917,925,930,939,944,949,958,963,971,976,1179,1184,1192,1215,1221,1226,1259,1264,1270,1275,1280,1285,1308,1313,1318,1324,1329,1342,1355,1375,1388,1409,1421,1441,1446,1452,1457,1462,1475,1480,1541,1546,1588,1593,1599,1604,1610,1615,1638,1643,1649,1695,1700,1724,1730,1740,1745,1763,1812,1817,1822,1827,1833,1838,1871,1876,1911,1916,1970,1975,1981,1986,1994,2017],{"type":39,"tag":40,"props":41,"children":42},"element","h1",{"id":4},[43],{"type":44,"value":45},"text","Skill Creator",{"type":39,"tag":47,"props":48,"children":49},"p",{},[50],{"type":44,"value":51},"This skill provides guidance for creating effective skills.",{"type":39,"tag":53,"props":54,"children":56},"h2",{"id":55},"about-skills",[57],{"type":44,"value":58},"About Skills",{"type":39,"tag":47,"props":60,"children":61},{},[62],{"type":44,"value":63},"Skills are modular, self-contained packages that extend Claude's capabilities by providing\nspecialized knowledge, workflows, and tools. Think of them as \"onboarding guides\" for specific\ndomains or tasks—they transform Claude from a general-purpose agent into a specialized agent\nequipped with procedural knowledge that no model can fully possess.",{"type":39,"tag":65,"props":66,"children":68},"h3",{"id":67},"what-skills-provide",[69],{"type":44,"value":70},"What Skills Provide",{"type":39,"tag":72,"props":73,"children":74},"ol",{},[75,81,86,91],{"type":39,"tag":76,"props":77,"children":78},"li",{},[79],{"type":44,"value":80},"Specialized workflows - Multi-step procedures for specific domains",{"type":39,"tag":76,"props":82,"children":83},{},[84],{"type":44,"value":85},"Tool integrations - Instructions for working with specific file formats or APIs",{"type":39,"tag":76,"props":87,"children":88},{},[89],{"type":44,"value":90},"Domain expertise - Company-specific knowledge, schemas, business logic",{"type":39,"tag":76,"props":92,"children":93},{},[94],{"type":44,"value":95},"Bundled resources - Scripts, references, and assets for complex and repetitive tasks",{"type":39,"tag":53,"props":97,"children":99},{"id":98},"core-principles",[100],{"type":44,"value":101},"Core Principles",{"type":39,"tag":65,"props":103,"children":105},{"id":104},"concise-is-key",[106],{"type":44,"value":107},"Concise is Key",{"type":39,"tag":47,"props":109,"children":110},{},[111],{"type":44,"value":112},"The context window is a public good. Skills share the context window with everything else Claude needs: system prompt, conversation history, other Skills' metadata, and the actual user request.",{"type":39,"tag":47,"props":114,"children":115},{},[116,122],{"type":39,"tag":117,"props":118,"children":119},"strong",{},[120],{"type":44,"value":121},"Default assumption: Claude is already very smart.",{"type":44,"value":123}," Only add context Claude doesn't already have. Challenge each piece of information: \"Does Claude really need this explanation?\" and \"Does this paragraph justify its token cost?\"",{"type":39,"tag":47,"props":125,"children":126},{},[127],{"type":44,"value":128},"Prefer concise examples over verbose explanations.",{"type":39,"tag":65,"props":130,"children":132},{"id":131},"set-appropriate-degrees-of-freedom",[133],{"type":44,"value":134},"Set Appropriate Degrees of Freedom",{"type":39,"tag":47,"props":136,"children":137},{},[138],{"type":44,"value":139},"Match the level of specificity to the task's fragility and variability:",{"type":39,"tag":47,"props":141,"children":142},{},[143,148],{"type":39,"tag":117,"props":144,"children":145},{},[146],{"type":44,"value":147},"High freedom (text-based instructions)",{"type":44,"value":149},": Use when multiple approaches are valid, decisions depend on context, or heuristics guide the approach.",{"type":39,"tag":47,"props":151,"children":152},{},[153,158],{"type":39,"tag":117,"props":154,"children":155},{},[156],{"type":44,"value":157},"Medium freedom (pseudocode or scripts with parameters)",{"type":44,"value":159},": Use when a preferred pattern exists, some variation is acceptable, or configuration affects behavior.",{"type":39,"tag":47,"props":161,"children":162},{},[163,168],{"type":39,"tag":117,"props":164,"children":165},{},[166],{"type":44,"value":167},"Low freedom (specific scripts, few parameters)",{"type":44,"value":169},": Use when operations are fragile and error-prone, consistency is critical, or a specific sequence must be followed.",{"type":39,"tag":47,"props":171,"children":172},{},[173],{"type":44,"value":174},"Think of Claude as exploring a path: a narrow bridge with cliffs needs specific guardrails (low freedom), while an open field allows many routes (high freedom).",{"type":39,"tag":65,"props":176,"children":178},{"id":177},"anatomy-of-a-skill",[179],{"type":44,"value":180},"Anatomy of a Skill",{"type":39,"tag":47,"props":182,"children":183},{},[184],{"type":44,"value":185},"Every skill consists of a required SKILL.md file and optional bundled resources:",{"type":39,"tag":187,"props":188,"children":192},"pre",{"className":189,"code":191,"language":44},[190],"language-text","skill-name\u002F\n├── SKILL.md (required)\n│   ├── YAML frontmatter metadata (required)\n│   │   ├── name: (required)\n│   │   └── description: (required)\n│   └── Markdown instructions (required)\n└── Bundled Resources (optional)\n    ├── scripts\u002F          - Executable code (Python\u002FBash\u002Fetc.)\n    ├── references\u002F       - Documentation intended to be loaded into context as needed\n    └── assets\u002F           - Files used in output (templates, icons, fonts, etc.)\n",[193],{"type":39,"tag":194,"props":195,"children":197},"code",{"__ignoreMap":196},"",[198],{"type":44,"value":191},{"type":39,"tag":200,"props":201,"children":203},"h4",{"id":202},"skillmd-required",[204],{"type":44,"value":205},"SKILL.md (required)",{"type":39,"tag":47,"props":207,"children":208},{},[209],{"type":44,"value":210},"Every SKILL.md consists of:",{"type":39,"tag":212,"props":213,"children":214},"ul",{},[215,241],{"type":39,"tag":76,"props":216,"children":217},{},[218,223,225,231,233,239],{"type":39,"tag":117,"props":219,"children":220},{},[221],{"type":44,"value":222},"Frontmatter",{"type":44,"value":224}," (YAML): Contains ",{"type":39,"tag":194,"props":226,"children":228},{"className":227},[],[229],{"type":44,"value":230},"name",{"type":44,"value":232}," and ",{"type":39,"tag":194,"props":234,"children":236},{"className":235},[],[237],{"type":44,"value":238},"description",{"type":44,"value":240}," fields. These are the only fields that Claude reads to determine when the skill gets used, thus it is very important to be clear and comprehensive in describing what the skill is, and when it should be used.",{"type":39,"tag":76,"props":242,"children":243},{},[244,249],{"type":39,"tag":117,"props":245,"children":246},{},[247],{"type":44,"value":248},"Body",{"type":44,"value":250}," (Markdown): Instructions and guidance for using the skill. Only loaded AFTER the skill triggers (if at all).",{"type":39,"tag":200,"props":252,"children":254},{"id":253},"bundled-resources-optional",[255],{"type":44,"value":256},"Bundled Resources (optional)",{"type":39,"tag":258,"props":259,"children":261},"h5",{"id":260},"scripts-scripts",[262,264,270],{"type":44,"value":263},"Scripts (",{"type":39,"tag":194,"props":265,"children":267},{"className":266},[],[268],{"type":44,"value":269},"scripts\u002F",{"type":44,"value":271},")",{"type":39,"tag":47,"props":273,"children":274},{},[275],{"type":44,"value":276},"Executable code (Python\u002FBash\u002Fetc.) for tasks that require deterministic reliability or are repeatedly rewritten.",{"type":39,"tag":212,"props":278,"children":279},{},[280,290,308,318],{"type":39,"tag":76,"props":281,"children":282},{},[283,288],{"type":39,"tag":117,"props":284,"children":285},{},[286],{"type":44,"value":287},"When to include",{"type":44,"value":289},": When the same code is being rewritten repeatedly or deterministic reliability is needed",{"type":39,"tag":76,"props":291,"children":292},{},[293,298,300,306],{"type":39,"tag":117,"props":294,"children":295},{},[296],{"type":44,"value":297},"Example",{"type":44,"value":299},": ",{"type":39,"tag":194,"props":301,"children":303},{"className":302},[],[304],{"type":44,"value":305},"scripts\u002Frotate_pdf.py",{"type":44,"value":307}," for PDF rotation tasks",{"type":39,"tag":76,"props":309,"children":310},{},[311,316],{"type":39,"tag":117,"props":312,"children":313},{},[314],{"type":44,"value":315},"Benefits",{"type":44,"value":317},": Token efficient, deterministic, may be executed without loading into context",{"type":39,"tag":76,"props":319,"children":320},{},[321,326],{"type":39,"tag":117,"props":322,"children":323},{},[324],{"type":44,"value":325},"Note",{"type":44,"value":327},": Scripts may still need to be read by Claude for patching or environment-specific adjustments",{"type":39,"tag":258,"props":329,"children":331},{"id":330},"references-references",[332,334,340],{"type":44,"value":333},"References (",{"type":39,"tag":194,"props":335,"children":337},{"className":336},[],[338],{"type":44,"value":339},"references\u002F",{"type":44,"value":271},{"type":39,"tag":47,"props":342,"children":343},{},[344],{"type":44,"value":345},"Documentation and reference material intended to be loaded as needed into context to inform Claude's process and thinking.",{"type":39,"tag":212,"props":347,"children":348},{},[349,358,399,409,418,428],{"type":39,"tag":76,"props":350,"children":351},{},[352,356],{"type":39,"tag":117,"props":353,"children":354},{},[355],{"type":44,"value":287},{"type":44,"value":357},": For documentation that Claude should reference while working",{"type":39,"tag":76,"props":359,"children":360},{},[361,366,367,373,375,381,383,389,391,397],{"type":39,"tag":117,"props":362,"children":363},{},[364],{"type":44,"value":365},"Examples",{"type":44,"value":299},{"type":39,"tag":194,"props":368,"children":370},{"className":369},[],[371],{"type":44,"value":372},"references\u002Ffinance.md",{"type":44,"value":374}," for financial schemas, ",{"type":39,"tag":194,"props":376,"children":378},{"className":377},[],[379],{"type":44,"value":380},"references\u002Fmnda.md",{"type":44,"value":382}," for company NDA template, ",{"type":39,"tag":194,"props":384,"children":386},{"className":385},[],[387],{"type":44,"value":388},"references\u002Fpolicies.md",{"type":44,"value":390}," for company policies, ",{"type":39,"tag":194,"props":392,"children":394},{"className":393},[],[395],{"type":44,"value":396},"references\u002Fapi_docs.md",{"type":44,"value":398}," for API specifications",{"type":39,"tag":76,"props":400,"children":401},{},[402,407],{"type":39,"tag":117,"props":403,"children":404},{},[405],{"type":44,"value":406},"Use cases",{"type":44,"value":408},": Database schemas, API documentation, domain knowledge, company policies, detailed workflow guides",{"type":39,"tag":76,"props":410,"children":411},{},[412,416],{"type":39,"tag":117,"props":413,"children":414},{},[415],{"type":44,"value":315},{"type":44,"value":417},": Keeps SKILL.md lean, loaded only when Claude determines it's needed",{"type":39,"tag":76,"props":419,"children":420},{},[421,426],{"type":39,"tag":117,"props":422,"children":423},{},[424],{"type":44,"value":425},"Best practice",{"type":44,"value":427},": If files are large (>10k words), include grep search patterns in SKILL.md",{"type":39,"tag":76,"props":429,"children":430},{},[431,436],{"type":39,"tag":117,"props":432,"children":433},{},[434],{"type":44,"value":435},"Avoid duplication",{"type":44,"value":437},": Information should live in either SKILL.md or references files, not both. Prefer references files for detailed information unless it's truly core to the skill—this keeps SKILL.md lean while making information discoverable without hogging the context window. Keep only essential procedural instructions and workflow guidance in SKILL.md; move detailed reference material, schemas, and examples to references files.",{"type":39,"tag":258,"props":439,"children":441},{"id":440},"assets-assets",[442,444,450],{"type":44,"value":443},"Assets (",{"type":39,"tag":194,"props":445,"children":447},{"className":446},[],[448],{"type":44,"value":449},"assets\u002F",{"type":44,"value":271},{"type":39,"tag":47,"props":452,"children":453},{},[454],{"type":44,"value":455},"Files not intended to be loaded into context, but rather used within the output Claude produces.",{"type":39,"tag":212,"props":457,"children":458},{},[459,468,508,517],{"type":39,"tag":76,"props":460,"children":461},{},[462,466],{"type":39,"tag":117,"props":463,"children":464},{},[465],{"type":44,"value":287},{"type":44,"value":467},": When the skill needs files that will be used in the final output",{"type":39,"tag":76,"props":469,"children":470},{},[471,475,476,482,484,490,492,498,500,506],{"type":39,"tag":117,"props":472,"children":473},{},[474],{"type":44,"value":365},{"type":44,"value":299},{"type":39,"tag":194,"props":477,"children":479},{"className":478},[],[480],{"type":44,"value":481},"assets\u002Flogo.png",{"type":44,"value":483}," for brand assets, ",{"type":39,"tag":194,"props":485,"children":487},{"className":486},[],[488],{"type":44,"value":489},"assets\u002Fslides.pptx",{"type":44,"value":491}," for PowerPoint templates, ",{"type":39,"tag":194,"props":493,"children":495},{"className":494},[],[496],{"type":44,"value":497},"assets\u002Ffrontend-template\u002F",{"type":44,"value":499}," for HTML\u002FReact boilerplate, ",{"type":39,"tag":194,"props":501,"children":503},{"className":502},[],[504],{"type":44,"value":505},"assets\u002Ffont.ttf",{"type":44,"value":507}," for typography",{"type":39,"tag":76,"props":509,"children":510},{},[511,515],{"type":39,"tag":117,"props":512,"children":513},{},[514],{"type":44,"value":406},{"type":44,"value":516},": Templates, images, icons, boilerplate code, fonts, sample documents that get copied or modified",{"type":39,"tag":76,"props":518,"children":519},{},[520,524],{"type":39,"tag":117,"props":521,"children":522},{},[523],{"type":44,"value":315},{"type":44,"value":525},": Separates output resources from documentation, enables Claude to use files without loading them into context",{"type":39,"tag":200,"props":527,"children":529},{"id":528},"what-to-not-include-in-a-skill",[530],{"type":44,"value":531},"What to Not Include in a Skill",{"type":39,"tag":47,"props":533,"children":534},{},[535],{"type":44,"value":536},"A skill should only contain essential files that directly support its functionality. Do NOT create extraneous documentation or auxiliary files, including:",{"type":39,"tag":212,"props":538,"children":539},{},[540,545,550,555,560],{"type":39,"tag":76,"props":541,"children":542},{},[543],{"type":44,"value":544},"README.md",{"type":39,"tag":76,"props":546,"children":547},{},[548],{"type":44,"value":549},"INSTALLATION_GUIDE.md",{"type":39,"tag":76,"props":551,"children":552},{},[553],{"type":44,"value":554},"QUICK_REFERENCE.md",{"type":39,"tag":76,"props":556,"children":557},{},[558],{"type":44,"value":559},"CHANGELOG.md",{"type":39,"tag":76,"props":561,"children":562},{},[563],{"type":44,"value":564},"etc.",{"type":39,"tag":47,"props":566,"children":567},{},[568],{"type":44,"value":569},"The skill should only contain the information needed for an AI agent to do the job at hand. It should not contain auxilary context about the process that went into creating it, setup and testing procedures, user-facing documentation, etc. Creating additional documentation files just adds clutter and confusion.",{"type":39,"tag":65,"props":571,"children":573},{"id":572},"progressive-disclosure-design-principle",[574],{"type":44,"value":575},"Progressive Disclosure Design Principle",{"type":39,"tag":47,"props":577,"children":578},{},[579],{"type":44,"value":580},"Skills use a three-level loading system to manage context efficiently:",{"type":39,"tag":72,"props":582,"children":583},{},[584,594,604],{"type":39,"tag":76,"props":585,"children":586},{},[587,592],{"type":39,"tag":117,"props":588,"children":589},{},[590],{"type":44,"value":591},"Metadata (name + description)",{"type":44,"value":593}," - Always in context (~100 words)",{"type":39,"tag":76,"props":595,"children":596},{},[597,602],{"type":39,"tag":117,"props":598,"children":599},{},[600],{"type":44,"value":601},"SKILL.md body",{"type":44,"value":603}," - When skill triggers (\u003C5k words)",{"type":39,"tag":76,"props":605,"children":606},{},[607,612],{"type":39,"tag":117,"props":608,"children":609},{},[610],{"type":44,"value":611},"Bundled resources",{"type":44,"value":613}," - As needed by Claude (Unlimited because scripts can be executed without reading into context window)",{"type":39,"tag":200,"props":615,"children":617},{"id":616},"progressive-disclosure-patterns",[618],{"type":44,"value":619},"Progressive Disclosure Patterns",{"type":39,"tag":47,"props":621,"children":622},{},[623],{"type":44,"value":624},"Keep SKILL.md body to the essentials and under 500 lines to minimize context bloat. Split content into separate files when approaching this limit. When splitting out content into other files, it is very important to reference them from SKILL.md and describe clearly when to read them, to ensure the reader of the skill knows they exist and when to use them.",{"type":39,"tag":47,"props":626,"children":627},{},[628,633],{"type":39,"tag":117,"props":629,"children":630},{},[631],{"type":44,"value":632},"Key principle:",{"type":44,"value":634}," When a skill supports multiple variations, frameworks, or options, keep only the core workflow and selection guidance in SKILL.md. Move variant-specific details (patterns, examples, configuration) into separate reference files.",{"type":39,"tag":47,"props":636,"children":637},{},[638],{"type":39,"tag":117,"props":639,"children":640},{},[641],{"type":44,"value":642},"Pattern 1: High-level guide with references",{"type":39,"tag":187,"props":644,"children":648},{"className":645,"code":646,"language":647,"meta":196,"style":196},"language-markdown shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","# PDF Processing\n\n## Quick start\n\nExtract text with pdfplumber:\n[code example]\n\n## Advanced features\n\n- **Form filling**: See [FORMS.md](FORMS.md) for complete guide\n- **API reference**: See [REFERENCE.md](REFERENCE.md) for all methods\n- **Examples**: See [EXAMPLES.md](EXAMPLES.md) for common patterns\n","markdown",[649],{"type":39,"tag":194,"props":650,"children":651},{"__ignoreMap":196},[652,670,680,694,702,712,721,729,742,750,811,862],{"type":39,"tag":653,"props":654,"children":657},"span",{"class":655,"line":656},"line",1,[658,664],{"type":39,"tag":653,"props":659,"children":661},{"style":660},"--shiki-light:#39ADB5;--shiki-default:#89DDFF;--shiki-dark:#89DDFF",[662],{"type":44,"value":663},"# ",{"type":39,"tag":653,"props":665,"children":667},{"style":666},"--shiki-light:#E2931D;--shiki-default:#FFCB6B;--shiki-dark:#FFCB6B",[668],{"type":44,"value":669},"PDF Processing\n",{"type":39,"tag":653,"props":671,"children":673},{"class":655,"line":672},2,[674],{"type":39,"tag":653,"props":675,"children":677},{"emptyLinePlaceholder":676},true,[678],{"type":44,"value":679},"\n",{"type":39,"tag":653,"props":681,"children":683},{"class":655,"line":682},3,[684,689],{"type":39,"tag":653,"props":685,"children":686},{"style":660},[687],{"type":44,"value":688},"## ",{"type":39,"tag":653,"props":690,"children":691},{"style":666},[692],{"type":44,"value":693},"Quick start\n",{"type":39,"tag":653,"props":695,"children":697},{"class":655,"line":696},4,[698],{"type":39,"tag":653,"props":699,"children":700},{"emptyLinePlaceholder":676},[701],{"type":44,"value":679},{"type":39,"tag":653,"props":703,"children":705},{"class":655,"line":704},5,[706],{"type":39,"tag":653,"props":707,"children":709},{"style":708},"--shiki-light:#90A4AE;--shiki-default:#EEFFFF;--shiki-dark:#BABED8",[710],{"type":44,"value":711},"Extract text with pdfplumber:\n",{"type":39,"tag":653,"props":713,"children":715},{"class":655,"line":714},6,[716],{"type":39,"tag":653,"props":717,"children":718},{"style":708},[719],{"type":44,"value":720},"[code example]\n",{"type":39,"tag":653,"props":722,"children":724},{"class":655,"line":723},7,[725],{"type":39,"tag":653,"props":726,"children":727},{"emptyLinePlaceholder":676},[728],{"type":44,"value":679},{"type":39,"tag":653,"props":730,"children":732},{"class":655,"line":731},8,[733,737],{"type":39,"tag":653,"props":734,"children":735},{"style":660},[736],{"type":44,"value":688},{"type":39,"tag":653,"props":738,"children":739},{"style":666},[740],{"type":44,"value":741},"Advanced features\n",{"type":39,"tag":653,"props":743,"children":745},{"class":655,"line":744},9,[746],{"type":39,"tag":653,"props":747,"children":748},{"emptyLinePlaceholder":676},[749],{"type":44,"value":679},{"type":39,"tag":653,"props":751,"children":753},{"class":655,"line":752},10,[754,759,765,771,776,781,786,792,797,802,806],{"type":39,"tag":653,"props":755,"children":756},{"style":660},[757],{"type":44,"value":758},"-",{"type":39,"tag":653,"props":760,"children":762},{"style":761},"--shiki-light:#39ADB5;--shiki-light-font-weight:bold;--shiki-default:#89DDFF;--shiki-default-font-weight:bold;--shiki-dark:#89DDFF;--shiki-dark-font-weight:bold",[763],{"type":44,"value":764}," **",{"type":39,"tag":653,"props":766,"children":768},{"style":767},"--shiki-light:#E53935;--shiki-light-font-weight:bold;--shiki-default:#F07178;--shiki-default-font-weight:bold;--shiki-dark:#F07178;--shiki-dark-font-weight:bold",[769],{"type":44,"value":770},"Form filling",{"type":39,"tag":653,"props":772,"children":773},{"style":761},[774],{"type":44,"value":775},"**",{"type":39,"tag":653,"props":777,"children":778},{"style":708},[779],{"type":44,"value":780},": See ",{"type":39,"tag":653,"props":782,"children":783},{"style":660},[784],{"type":44,"value":785},"[",{"type":39,"tag":653,"props":787,"children":789},{"style":788},"--shiki-light:#91B859;--shiki-default:#C3E88D;--shiki-dark:#C3E88D",[790],{"type":44,"value":791},"FORMS.md",{"type":39,"tag":653,"props":793,"children":794},{"style":660},[795],{"type":44,"value":796},"](",{"type":39,"tag":653,"props":798,"children":800},{"style":799},"--shiki-light:#E53935;--shiki-light-text-decoration:underline;--shiki-default:#F07178;--shiki-default-text-decoration:underline;--shiki-dark:#F07178;--shiki-dark-text-decoration:underline",[801],{"type":44,"value":791},{"type":39,"tag":653,"props":803,"children":804},{"style":660},[805],{"type":44,"value":271},{"type":39,"tag":653,"props":807,"children":808},{"style":708},[809],{"type":44,"value":810}," for complete guide\n",{"type":39,"tag":653,"props":812,"children":814},{"class":655,"line":813},11,[815,819,823,828,832,836,840,845,849,853,857],{"type":39,"tag":653,"props":816,"children":817},{"style":660},[818],{"type":44,"value":758},{"type":39,"tag":653,"props":820,"children":821},{"style":761},[822],{"type":44,"value":764},{"type":39,"tag":653,"props":824,"children":825},{"style":767},[826],{"type":44,"value":827},"API reference",{"type":39,"tag":653,"props":829,"children":830},{"style":761},[831],{"type":44,"value":775},{"type":39,"tag":653,"props":833,"children":834},{"style":708},[835],{"type":44,"value":780},{"type":39,"tag":653,"props":837,"children":838},{"style":660},[839],{"type":44,"value":785},{"type":39,"tag":653,"props":841,"children":842},{"style":788},[843],{"type":44,"value":844},"REFERENCE.md",{"type":39,"tag":653,"props":846,"children":847},{"style":660},[848],{"type":44,"value":796},{"type":39,"tag":653,"props":850,"children":851},{"style":799},[852],{"type":44,"value":844},{"type":39,"tag":653,"props":854,"children":855},{"style":660},[856],{"type":44,"value":271},{"type":39,"tag":653,"props":858,"children":859},{"style":708},[860],{"type":44,"value":861}," for all methods\n",{"type":39,"tag":653,"props":863,"children":865},{"class":655,"line":864},12,[866,870,874,878,882,886,890,895,899,903,907],{"type":39,"tag":653,"props":867,"children":868},{"style":660},[869],{"type":44,"value":758},{"type":39,"tag":653,"props":871,"children":872},{"style":761},[873],{"type":44,"value":764},{"type":39,"tag":653,"props":875,"children":876},{"style":767},[877],{"type":44,"value":365},{"type":39,"tag":653,"props":879,"children":880},{"style":761},[881],{"type":44,"value":775},{"type":39,"tag":653,"props":883,"children":884},{"style":708},[885],{"type":44,"value":780},{"type":39,"tag":653,"props":887,"children":888},{"style":660},[889],{"type":44,"value":785},{"type":39,"tag":653,"props":891,"children":892},{"style":788},[893],{"type":44,"value":894},"EXAMPLES.md",{"type":39,"tag":653,"props":896,"children":897},{"style":660},[898],{"type":44,"value":796},{"type":39,"tag":653,"props":900,"children":901},{"style":799},[902],{"type":44,"value":894},{"type":39,"tag":653,"props":904,"children":905},{"style":660},[906],{"type":44,"value":271},{"type":39,"tag":653,"props":908,"children":909},{"style":708},[910],{"type":44,"value":911}," for common patterns\n",{"type":39,"tag":47,"props":913,"children":914},{},[915],{"type":44,"value":916},"Claude loads FORMS.md, REFERENCE.md, or EXAMPLES.md only when needed.",{"type":39,"tag":47,"props":918,"children":919},{},[920],{"type":39,"tag":117,"props":921,"children":922},{},[923],{"type":44,"value":924},"Pattern 2: Domain-specific organization",{"type":39,"tag":47,"props":926,"children":927},{},[928],{"type":44,"value":929},"For Skills with multiple domains, organize content by domain to avoid loading irrelevant context:",{"type":39,"tag":187,"props":931,"children":934},{"className":932,"code":933,"language":44},[190],"bigquery-skill\u002F\n├── SKILL.md (overview and navigation)\n└── reference\u002F\n    ├── finance.md (revenue, billing metrics)\n    ├── sales.md (opportunities, pipeline)\n    ├── product.md (API usage, features)\n    └── marketing.md (campaigns, attribution)\n",[935],{"type":39,"tag":194,"props":936,"children":937},{"__ignoreMap":196},[938],{"type":44,"value":933},{"type":39,"tag":47,"props":940,"children":941},{},[942],{"type":44,"value":943},"When a user asks about sales metrics, Claude only reads sales.md.",{"type":39,"tag":47,"props":945,"children":946},{},[947],{"type":44,"value":948},"Similarly, for skills supporting multiple frameworks or variants, organize by variant:",{"type":39,"tag":187,"props":950,"children":953},{"className":951,"code":952,"language":44},[190],"cloud-deploy\u002F\n├── SKILL.md (workflow + provider selection)\n└── references\u002F\n    ├── aws.md (AWS deployment patterns)\n    ├── gcp.md (GCP deployment patterns)\n    └── azure.md (Azure deployment patterns)\n",[954],{"type":39,"tag":194,"props":955,"children":956},{"__ignoreMap":196},[957],{"type":44,"value":952},{"type":39,"tag":47,"props":959,"children":960},{},[961],{"type":44,"value":962},"When the user chooses AWS, Claude only reads aws.md.",{"type":39,"tag":47,"props":964,"children":965},{},[966],{"type":39,"tag":117,"props":967,"children":968},{},[969],{"type":44,"value":970},"Pattern 3: Conditional details",{"type":39,"tag":47,"props":972,"children":973},{},[974],{"type":44,"value":975},"Show basic content, link to advanced content:",{"type":39,"tag":187,"props":977,"children":979},{"className":645,"code":978,"language":647,"meta":196,"style":196},"# DOCX Processing\n\n## Creating documents\n\nUse docx-js for new documents. See [DOCX-JS.md](DOCX-JS.md).\n\n## Editing documents\n\nFor simple edits, modify the XML directly.\n\n**For tracked changes**: See [REDLINING.md](REDLINING.md)\n**For OOXML details**: See [OOXML.md](OOXML.md)\n",[980],{"type":39,"tag":194,"props":981,"children":982},{"__ignoreMap":196},[983,995,1002,1014,1021,1055,1062,1074,1081,1089,1096,1138],{"type":39,"tag":653,"props":984,"children":985},{"class":655,"line":656},[986,990],{"type":39,"tag":653,"props":987,"children":988},{"style":660},[989],{"type":44,"value":663},{"type":39,"tag":653,"props":991,"children":992},{"style":666},[993],{"type":44,"value":994},"DOCX Processing\n",{"type":39,"tag":653,"props":996,"children":997},{"class":655,"line":672},[998],{"type":39,"tag":653,"props":999,"children":1000},{"emptyLinePlaceholder":676},[1001],{"type":44,"value":679},{"type":39,"tag":653,"props":1003,"children":1004},{"class":655,"line":682},[1005,1009],{"type":39,"tag":653,"props":1006,"children":1007},{"style":660},[1008],{"type":44,"value":688},{"type":39,"tag":653,"props":1010,"children":1011},{"style":666},[1012],{"type":44,"value":1013},"Creating documents\n",{"type":39,"tag":653,"props":1015,"children":1016},{"class":655,"line":696},[1017],{"type":39,"tag":653,"props":1018,"children":1019},{"emptyLinePlaceholder":676},[1020],{"type":44,"value":679},{"type":39,"tag":653,"props":1022,"children":1023},{"class":655,"line":704},[1024,1029,1033,1038,1042,1046,1050],{"type":39,"tag":653,"props":1025,"children":1026},{"style":708},[1027],{"type":44,"value":1028},"Use docx-js for new documents. See ",{"type":39,"tag":653,"props":1030,"children":1031},{"style":660},[1032],{"type":44,"value":785},{"type":39,"tag":653,"props":1034,"children":1035},{"style":788},[1036],{"type":44,"value":1037},"DOCX-JS.md",{"type":39,"tag":653,"props":1039,"children":1040},{"style":660},[1041],{"type":44,"value":796},{"type":39,"tag":653,"props":1043,"children":1044},{"style":799},[1045],{"type":44,"value":1037},{"type":39,"tag":653,"props":1047,"children":1048},{"style":660},[1049],{"type":44,"value":271},{"type":39,"tag":653,"props":1051,"children":1052},{"style":708},[1053],{"type":44,"value":1054},".\n",{"type":39,"tag":653,"props":1056,"children":1057},{"class":655,"line":714},[1058],{"type":39,"tag":653,"props":1059,"children":1060},{"emptyLinePlaceholder":676},[1061],{"type":44,"value":679},{"type":39,"tag":653,"props":1063,"children":1064},{"class":655,"line":723},[1065,1069],{"type":39,"tag":653,"props":1066,"children":1067},{"style":660},[1068],{"type":44,"value":688},{"type":39,"tag":653,"props":1070,"children":1071},{"style":666},[1072],{"type":44,"value":1073},"Editing documents\n",{"type":39,"tag":653,"props":1075,"children":1076},{"class":655,"line":731},[1077],{"type":39,"tag":653,"props":1078,"children":1079},{"emptyLinePlaceholder":676},[1080],{"type":44,"value":679},{"type":39,"tag":653,"props":1082,"children":1083},{"class":655,"line":744},[1084],{"type":39,"tag":653,"props":1085,"children":1086},{"style":708},[1087],{"type":44,"value":1088},"For simple edits, modify the XML directly.\n",{"type":39,"tag":653,"props":1090,"children":1091},{"class":655,"line":752},[1092],{"type":39,"tag":653,"props":1093,"children":1094},{"emptyLinePlaceholder":676},[1095],{"type":44,"value":679},{"type":39,"tag":653,"props":1097,"children":1098},{"class":655,"line":813},[1099,1103,1108,1112,1116,1120,1125,1129,1133],{"type":39,"tag":653,"props":1100,"children":1101},{"style":761},[1102],{"type":44,"value":775},{"type":39,"tag":653,"props":1104,"children":1105},{"style":767},[1106],{"type":44,"value":1107},"For tracked changes",{"type":39,"tag":653,"props":1109,"children":1110},{"style":761},[1111],{"type":44,"value":775},{"type":39,"tag":653,"props":1113,"children":1114},{"style":708},[1115],{"type":44,"value":780},{"type":39,"tag":653,"props":1117,"children":1118},{"style":660},[1119],{"type":44,"value":785},{"type":39,"tag":653,"props":1121,"children":1122},{"style":788},[1123],{"type":44,"value":1124},"REDLINING.md",{"type":39,"tag":653,"props":1126,"children":1127},{"style":660},[1128],{"type":44,"value":796},{"type":39,"tag":653,"props":1130,"children":1131},{"style":799},[1132],{"type":44,"value":1124},{"type":39,"tag":653,"props":1134,"children":1135},{"style":660},[1136],{"type":44,"value":1137},")\n",{"type":39,"tag":653,"props":1139,"children":1140},{"class":655,"line":864},[1141,1145,1150,1154,1158,1162,1167,1171,1175],{"type":39,"tag":653,"props":1142,"children":1143},{"style":761},[1144],{"type":44,"value":775},{"type":39,"tag":653,"props":1146,"children":1147},{"style":767},[1148],{"type":44,"value":1149},"For OOXML details",{"type":39,"tag":653,"props":1151,"children":1152},{"style":761},[1153],{"type":44,"value":775},{"type":39,"tag":653,"props":1155,"children":1156},{"style":708},[1157],{"type":44,"value":780},{"type":39,"tag":653,"props":1159,"children":1160},{"style":660},[1161],{"type":44,"value":785},{"type":39,"tag":653,"props":1163,"children":1164},{"style":788},[1165],{"type":44,"value":1166},"OOXML.md",{"type":39,"tag":653,"props":1168,"children":1169},{"style":660},[1170],{"type":44,"value":796},{"type":39,"tag":653,"props":1172,"children":1173},{"style":799},[1174],{"type":44,"value":1166},{"type":39,"tag":653,"props":1176,"children":1177},{"style":660},[1178],{"type":44,"value":1137},{"type":39,"tag":47,"props":1180,"children":1181},{},[1182],{"type":44,"value":1183},"Claude reads REDLINING.md or OOXML.md only when the user needs those features.",{"type":39,"tag":47,"props":1185,"children":1186},{},[1187],{"type":39,"tag":117,"props":1188,"children":1189},{},[1190],{"type":44,"value":1191},"Important guidelines:",{"type":39,"tag":212,"props":1193,"children":1194},{},[1195,1205],{"type":39,"tag":76,"props":1196,"children":1197},{},[1198,1203],{"type":39,"tag":117,"props":1199,"children":1200},{},[1201],{"type":44,"value":1202},"Avoid deeply nested references",{"type":44,"value":1204}," - Keep references one level deep from SKILL.md. All reference files should link directly from SKILL.md.",{"type":39,"tag":76,"props":1206,"children":1207},{},[1208,1213],{"type":39,"tag":117,"props":1209,"children":1210},{},[1211],{"type":44,"value":1212},"Structure longer reference files",{"type":44,"value":1214}," - For files longer than 100 lines, include a table of contents at the top so Claude can see the full scope when previewing.",{"type":39,"tag":53,"props":1216,"children":1218},{"id":1217},"skill-creation-process",[1219],{"type":44,"value":1220},"Skill Creation Process",{"type":39,"tag":47,"props":1222,"children":1223},{},[1224],{"type":44,"value":1225},"Skill creation involves these steps:",{"type":39,"tag":72,"props":1227,"children":1228},{},[1229,1234,1239,1244,1249,1254],{"type":39,"tag":76,"props":1230,"children":1231},{},[1232],{"type":44,"value":1233},"Understand the skill with concrete examples",{"type":39,"tag":76,"props":1235,"children":1236},{},[1237],{"type":44,"value":1238},"Plan reusable skill contents (scripts, references, assets)",{"type":39,"tag":76,"props":1240,"children":1241},{},[1242],{"type":44,"value":1243},"Initialize the skill (run init_skill.py)",{"type":39,"tag":76,"props":1245,"children":1246},{},[1247],{"type":44,"value":1248},"Edit the skill (implement resources and write SKILL.md)",{"type":39,"tag":76,"props":1250,"children":1251},{},[1252],{"type":44,"value":1253},"Package the skill (run package_skill.py)",{"type":39,"tag":76,"props":1255,"children":1256},{},[1257],{"type":44,"value":1258},"Iterate based on real usage",{"type":39,"tag":47,"props":1260,"children":1261},{},[1262],{"type":44,"value":1263},"Follow these steps in order, skipping only if there is a clear reason why they are not applicable.",{"type":39,"tag":65,"props":1265,"children":1267},{"id":1266},"step-1-understanding-the-skill-with-concrete-examples",[1268],{"type":44,"value":1269},"Step 1: Understanding the Skill with Concrete Examples",{"type":39,"tag":47,"props":1271,"children":1272},{},[1273],{"type":44,"value":1274},"Skip this step only when the skill's usage patterns are already clearly understood. It remains valuable even when working with an existing skill.",{"type":39,"tag":47,"props":1276,"children":1277},{},[1278],{"type":44,"value":1279},"To create an effective skill, clearly understand concrete examples of how the skill will be used. This understanding can come from either direct user examples or generated examples that are validated with user feedback.",{"type":39,"tag":47,"props":1281,"children":1282},{},[1283],{"type":44,"value":1284},"For example, when building an image-editor skill, relevant questions include:",{"type":39,"tag":212,"props":1286,"children":1287},{},[1288,1293,1298,1303],{"type":39,"tag":76,"props":1289,"children":1290},{},[1291],{"type":44,"value":1292},"\"What functionality should the image-editor skill support? Editing, rotating, anything else?\"",{"type":39,"tag":76,"props":1294,"children":1295},{},[1296],{"type":44,"value":1297},"\"Can you give some examples of how this skill would be used?\"",{"type":39,"tag":76,"props":1299,"children":1300},{},[1301],{"type":44,"value":1302},"\"I can imagine users asking for things like 'Remove the red-eye from this image' or 'Rotate this image'. Are there other ways you imagine this skill being used?\"",{"type":39,"tag":76,"props":1304,"children":1305},{},[1306],{"type":44,"value":1307},"\"What would a user say that should trigger this skill?\"",{"type":39,"tag":47,"props":1309,"children":1310},{},[1311],{"type":44,"value":1312},"To avoid overwhelming users, avoid asking too many questions in a single message. Start with the most important questions and follow up as needed for better effectiveness.",{"type":39,"tag":47,"props":1314,"children":1315},{},[1316],{"type":44,"value":1317},"Conclude this step when there is a clear sense of the functionality the skill should support.",{"type":39,"tag":65,"props":1319,"children":1321},{"id":1320},"step-2-planning-the-reusable-skill-contents",[1322],{"type":44,"value":1323},"Step 2: Planning the Reusable Skill Contents",{"type":39,"tag":47,"props":1325,"children":1326},{},[1327],{"type":44,"value":1328},"To turn concrete examples into an effective skill, analyze each example by:",{"type":39,"tag":72,"props":1330,"children":1331},{},[1332,1337],{"type":39,"tag":76,"props":1333,"children":1334},{},[1335],{"type":44,"value":1336},"Considering how to execute on the example from scratch",{"type":39,"tag":76,"props":1338,"children":1339},{},[1340],{"type":44,"value":1341},"Identifying what scripts, references, and assets would be helpful when executing these workflows repeatedly",{"type":39,"tag":47,"props":1343,"children":1344},{},[1345,1347,1353],{"type":44,"value":1346},"Example: When building a ",{"type":39,"tag":194,"props":1348,"children":1350},{"className":1349},[],[1351],{"type":44,"value":1352},"pdf-editor",{"type":44,"value":1354}," skill to handle queries like \"Help me rotate this PDF,\" the analysis shows:",{"type":39,"tag":72,"props":1356,"children":1357},{},[1358,1363],{"type":39,"tag":76,"props":1359,"children":1360},{},[1361],{"type":44,"value":1362},"Rotating a PDF requires re-writing the same code each time",{"type":39,"tag":76,"props":1364,"children":1365},{},[1366,1368,1373],{"type":44,"value":1367},"A ",{"type":39,"tag":194,"props":1369,"children":1371},{"className":1370},[],[1372],{"type":44,"value":305},{"type":44,"value":1374}," script would be helpful to store in the skill",{"type":39,"tag":47,"props":1376,"children":1377},{},[1378,1380,1386],{"type":44,"value":1379},"Example: When designing a ",{"type":39,"tag":194,"props":1381,"children":1383},{"className":1382},[],[1384],{"type":44,"value":1385},"frontend-webapp-builder",{"type":44,"value":1387}," skill for queries like \"Build me a todo app\" or \"Build me a dashboard to track my steps,\" the analysis shows:",{"type":39,"tag":72,"props":1389,"children":1390},{},[1391,1396],{"type":39,"tag":76,"props":1392,"children":1393},{},[1394],{"type":44,"value":1395},"Writing a frontend webapp requires the same boilerplate HTML\u002FReact each time",{"type":39,"tag":76,"props":1397,"children":1398},{},[1399,1401,1407],{"type":44,"value":1400},"An ",{"type":39,"tag":194,"props":1402,"children":1404},{"className":1403},[],[1405],{"type":44,"value":1406},"assets\u002Fhello-world\u002F",{"type":44,"value":1408}," template containing the boilerplate HTML\u002FReact project files would be helpful to store in the skill",{"type":39,"tag":47,"props":1410,"children":1411},{},[1412,1413,1419],{"type":44,"value":1346},{"type":39,"tag":194,"props":1414,"children":1416},{"className":1415},[],[1417],{"type":44,"value":1418},"big-query",{"type":44,"value":1420}," skill to handle queries like \"How many users have logged in today?\" the analysis shows:",{"type":39,"tag":72,"props":1422,"children":1423},{},[1424,1429],{"type":39,"tag":76,"props":1425,"children":1426},{},[1427],{"type":44,"value":1428},"Querying BigQuery requires re-discovering the table schemas and relationships each time",{"type":39,"tag":76,"props":1430,"children":1431},{},[1432,1433,1439],{"type":44,"value":1367},{"type":39,"tag":194,"props":1434,"children":1436},{"className":1435},[],[1437],{"type":44,"value":1438},"references\u002Fschema.md",{"type":44,"value":1440}," file documenting the table schemas would be helpful to store in the skill",{"type":39,"tag":47,"props":1442,"children":1443},{},[1444],{"type":44,"value":1445},"To establish the skill's contents, analyze each concrete example to create a list of the reusable resources to include: scripts, references, and assets.",{"type":39,"tag":65,"props":1447,"children":1449},{"id":1448},"step-3-initializing-the-skill",[1450],{"type":44,"value":1451},"Step 3: Initializing the Skill",{"type":39,"tag":47,"props":1453,"children":1454},{},[1455],{"type":44,"value":1456},"At this point, it is time to actually create the skill.",{"type":39,"tag":47,"props":1458,"children":1459},{},[1460],{"type":44,"value":1461},"Skip this step only if the skill being developed already exists, and iteration or packaging is needed. In this case, continue to the next step.",{"type":39,"tag":47,"props":1463,"children":1464},{},[1465,1467,1473],{"type":44,"value":1466},"When creating a new skill from scratch, always run the ",{"type":39,"tag":194,"props":1468,"children":1470},{"className":1469},[],[1471],{"type":44,"value":1472},"init_skill.py",{"type":44,"value":1474}," script. The script conveniently generates a new template skill directory that automatically includes everything a skill requires, making the skill creation process much more efficient and reliable.",{"type":39,"tag":47,"props":1476,"children":1477},{},[1478],{"type":44,"value":1479},"Usage:",{"type":39,"tag":187,"props":1481,"children":1485},{"className":1482,"code":1483,"language":1484,"meta":196,"style":196},"language-bash shiki shiki-themes material-theme-lighter material-theme material-theme-palenight","scripts\u002Finit_skill.py \u003Cskill-name> --path \u003Coutput-directory>\n","bash",[1486],{"type":39,"tag":194,"props":1487,"children":1488},{"__ignoreMap":196},[1489],{"type":39,"tag":653,"props":1490,"children":1491},{"class":655,"line":656},[1492,1497,1502,1507,1512,1517,1522,1526,1531,1536],{"type":39,"tag":653,"props":1493,"children":1494},{"style":666},[1495],{"type":44,"value":1496},"scripts\u002Finit_skill.py",{"type":39,"tag":653,"props":1498,"children":1499},{"style":660},[1500],{"type":44,"value":1501}," \u003C",{"type":39,"tag":653,"props":1503,"children":1504},{"style":788},[1505],{"type":44,"value":1506},"skill-nam",{"type":39,"tag":653,"props":1508,"children":1509},{"style":708},[1510],{"type":44,"value":1511},"e",{"type":39,"tag":653,"props":1513,"children":1514},{"style":660},[1515],{"type":44,"value":1516},">",{"type":39,"tag":653,"props":1518,"children":1519},{"style":788},[1520],{"type":44,"value":1521}," --path",{"type":39,"tag":653,"props":1523,"children":1524},{"style":660},[1525],{"type":44,"value":1501},{"type":39,"tag":653,"props":1527,"children":1528},{"style":788},[1529],{"type":44,"value":1530},"output-director",{"type":39,"tag":653,"props":1532,"children":1533},{"style":708},[1534],{"type":44,"value":1535},"y",{"type":39,"tag":653,"props":1537,"children":1538},{"style":660},[1539],{"type":44,"value":1540},">\n",{"type":39,"tag":47,"props":1542,"children":1543},{},[1544],{"type":44,"value":1545},"The script:",{"type":39,"tag":212,"props":1547,"children":1548},{},[1549,1554,1559,1583],{"type":39,"tag":76,"props":1550,"children":1551},{},[1552],{"type":44,"value":1553},"Creates the skill directory at the specified path",{"type":39,"tag":76,"props":1555,"children":1556},{},[1557],{"type":44,"value":1558},"Generates a SKILL.md template with proper frontmatter and TODO placeholders",{"type":39,"tag":76,"props":1560,"children":1561},{},[1562,1564,1569,1571,1576,1578],{"type":44,"value":1563},"Creates example resource directories: ",{"type":39,"tag":194,"props":1565,"children":1567},{"className":1566},[],[1568],{"type":44,"value":269},{"type":44,"value":1570},", ",{"type":39,"tag":194,"props":1572,"children":1574},{"className":1573},[],[1575],{"type":44,"value":339},{"type":44,"value":1577},", and ",{"type":39,"tag":194,"props":1579,"children":1581},{"className":1580},[],[1582],{"type":44,"value":449},{"type":39,"tag":76,"props":1584,"children":1585},{},[1586],{"type":44,"value":1587},"Adds example files in each directory that can be customized or deleted",{"type":39,"tag":47,"props":1589,"children":1590},{},[1591],{"type":44,"value":1592},"After initialization, customize or remove the generated SKILL.md and example files as needed.",{"type":39,"tag":65,"props":1594,"children":1596},{"id":1595},"step-4-edit-the-skill",[1597],{"type":44,"value":1598},"Step 4: Edit the Skill",{"type":39,"tag":47,"props":1600,"children":1601},{},[1602],{"type":44,"value":1603},"When editing the (newly-generated or existing) skill, remember that the skill is being created for another instance of Claude to use. Include information that would be beneficial and non-obvious to Claude. Consider what procedural knowledge, domain-specific details, or reusable assets would help another Claude instance execute these tasks more effectively.",{"type":39,"tag":200,"props":1605,"children":1607},{"id":1606},"learn-proven-design-patterns",[1608],{"type":44,"value":1609},"Learn Proven Design Patterns",{"type":39,"tag":47,"props":1611,"children":1612},{},[1613],{"type":44,"value":1614},"Consult these helpful guides based on your skill's needs:",{"type":39,"tag":212,"props":1616,"children":1617},{},[1618,1628],{"type":39,"tag":76,"props":1619,"children":1620},{},[1621,1626],{"type":39,"tag":117,"props":1622,"children":1623},{},[1624],{"type":44,"value":1625},"Multi-step processes",{"type":44,"value":1627},": See references\u002Fworkflows.md for sequential workflows and conditional logic",{"type":39,"tag":76,"props":1629,"children":1630},{},[1631,1636],{"type":39,"tag":117,"props":1632,"children":1633},{},[1634],{"type":44,"value":1635},"Specific output formats or quality standards",{"type":44,"value":1637},": See references\u002Foutput-patterns.md for template and example patterns",{"type":39,"tag":47,"props":1639,"children":1640},{},[1641],{"type":44,"value":1642},"These files contain established best practices for effective skill design.",{"type":39,"tag":200,"props":1644,"children":1646},{"id":1645},"start-with-reusable-skill-contents",[1647],{"type":44,"value":1648},"Start with Reusable Skill Contents",{"type":39,"tag":47,"props":1650,"children":1651},{},[1652,1654,1659,1660,1665,1666,1671,1673,1679,1681,1686,1688,1693],{"type":44,"value":1653},"To begin implementation, start with the reusable resources identified above: ",{"type":39,"tag":194,"props":1655,"children":1657},{"className":1656},[],[1658],{"type":44,"value":269},{"type":44,"value":1570},{"type":39,"tag":194,"props":1661,"children":1663},{"className":1662},[],[1664],{"type":44,"value":339},{"type":44,"value":1577},{"type":39,"tag":194,"props":1667,"children":1669},{"className":1668},[],[1670],{"type":44,"value":449},{"type":44,"value":1672}," files. Note that this step may require user input. For example, when implementing a ",{"type":39,"tag":194,"props":1674,"children":1676},{"className":1675},[],[1677],{"type":44,"value":1678},"brand-guidelines",{"type":44,"value":1680}," skill, the user may need to provide brand assets or templates to store in ",{"type":39,"tag":194,"props":1682,"children":1684},{"className":1683},[],[1685],{"type":44,"value":449},{"type":44,"value":1687},", or documentation to store in ",{"type":39,"tag":194,"props":1689,"children":1691},{"className":1690},[],[1692],{"type":44,"value":339},{"type":44,"value":1694},".",{"type":39,"tag":47,"props":1696,"children":1697},{},[1698],{"type":44,"value":1699},"Added scripts must be tested by actually running them to ensure there are no bugs and that the output matches what is expected. If there are many similar scripts, only a representative sample needs to be tested to ensure confidence that they all work while balancing time to completion.",{"type":39,"tag":47,"props":1701,"children":1702},{},[1703,1705,1710,1711,1716,1717,1722],{"type":44,"value":1704},"Any example files and directories not needed for the skill should be deleted. The initialization script creates example files in ",{"type":39,"tag":194,"props":1706,"children":1708},{"className":1707},[],[1709],{"type":44,"value":269},{"type":44,"value":1570},{"type":39,"tag":194,"props":1712,"children":1714},{"className":1713},[],[1715],{"type":44,"value":339},{"type":44,"value":1577},{"type":39,"tag":194,"props":1718,"children":1720},{"className":1719},[],[1721],{"type":44,"value":449},{"type":44,"value":1723}," to demonstrate structure, but most skills won't need all of them.",{"type":39,"tag":200,"props":1725,"children":1727},{"id":1726},"update-skillmd",[1728],{"type":44,"value":1729},"Update SKILL.md",{"type":39,"tag":47,"props":1731,"children":1732},{},[1733,1738],{"type":39,"tag":117,"props":1734,"children":1735},{},[1736],{"type":44,"value":1737},"Writing Guidelines:",{"type":44,"value":1739}," Always use imperative\u002Finfinitive form.",{"type":39,"tag":258,"props":1741,"children":1743},{"id":1742},"frontmatter",[1744],{"type":44,"value":222},{"type":39,"tag":47,"props":1746,"children":1747},{},[1748,1750,1755,1756,1761],{"type":44,"value":1749},"Write the YAML frontmatter with ",{"type":39,"tag":194,"props":1751,"children":1753},{"className":1752},[],[1754],{"type":44,"value":230},{"type":44,"value":232},{"type":39,"tag":194,"props":1757,"children":1759},{"className":1758},[],[1760],{"type":44,"value":238},{"type":44,"value":1762},":",{"type":39,"tag":212,"props":1764,"children":1765},{},[1766,1776],{"type":39,"tag":76,"props":1767,"children":1768},{},[1769,1774],{"type":39,"tag":194,"props":1770,"children":1772},{"className":1771},[],[1773],{"type":44,"value":230},{"type":44,"value":1775},": The skill name",{"type":39,"tag":76,"props":1777,"children":1778},{},[1779,1784,1786],{"type":39,"tag":194,"props":1780,"children":1782},{"className":1781},[],[1783],{"type":44,"value":238},{"type":44,"value":1785},": This is the primary triggering mechanism for your skill, and helps Claude understand when to use the skill.\n",{"type":39,"tag":212,"props":1787,"children":1788},{},[1789,1794,1799],{"type":39,"tag":76,"props":1790,"children":1791},{},[1792],{"type":44,"value":1793},"Include both what the Skill does and specific triggers\u002Fcontexts for when to use it.",{"type":39,"tag":76,"props":1795,"children":1796},{},[1797],{"type":44,"value":1798},"Include all \"when to use\" information here - Not in the body. The body is only loaded after triggering, so \"When to Use This Skill\" sections in the body are not helpful to Claude.",{"type":39,"tag":76,"props":1800,"children":1801},{},[1802,1804,1810],{"type":44,"value":1803},"Example description for a ",{"type":39,"tag":194,"props":1805,"children":1807},{"className":1806},[],[1808],{"type":44,"value":1809},"docx",{"type":44,"value":1811}," skill: \"Comprehensive document creation, editing, and analysis with support for tracked changes, comments, formatting preservation, and text extraction. Use when Claude needs to work with professional documents (.docx files) for: (1) Creating new documents, (2) Modifying or editing content, (3) Working with tracked changes, (4) Adding comments, or any other document tasks\"",{"type":39,"tag":47,"props":1813,"children":1814},{},[1815],{"type":44,"value":1816},"Do not include any other fields in YAML frontmatter.",{"type":39,"tag":258,"props":1818,"children":1820},{"id":1819},"body",[1821],{"type":44,"value":248},{"type":39,"tag":47,"props":1823,"children":1824},{},[1825],{"type":44,"value":1826},"Write instructions for using the skill and its bundled resources.",{"type":39,"tag":65,"props":1828,"children":1830},{"id":1829},"step-5-packaging-a-skill",[1831],{"type":44,"value":1832},"Step 5: Packaging a Skill",{"type":39,"tag":47,"props":1834,"children":1835},{},[1836],{"type":44,"value":1837},"Once development of the skill is complete, it must be packaged into a distributable .skill file that gets shared with the user. The packaging process automatically validates the skill first to ensure it meets all requirements:",{"type":39,"tag":187,"props":1839,"children":1841},{"className":1482,"code":1840,"language":1484,"meta":196,"style":196},"scripts\u002Fpackage_skill.py \u003Cpath\u002Fto\u002Fskill-folder>\n",[1842],{"type":39,"tag":194,"props":1843,"children":1844},{"__ignoreMap":196},[1845],{"type":39,"tag":653,"props":1846,"children":1847},{"class":655,"line":656},[1848,1853,1857,1862,1867],{"type":39,"tag":653,"props":1849,"children":1850},{"style":666},[1851],{"type":44,"value":1852},"scripts\u002Fpackage_skill.py",{"type":39,"tag":653,"props":1854,"children":1855},{"style":660},[1856],{"type":44,"value":1501},{"type":39,"tag":653,"props":1858,"children":1859},{"style":788},[1860],{"type":44,"value":1861},"path\u002Fto\u002Fskill-folde",{"type":39,"tag":653,"props":1863,"children":1864},{"style":708},[1865],{"type":44,"value":1866},"r",{"type":39,"tag":653,"props":1868,"children":1869},{"style":660},[1870],{"type":44,"value":1540},{"type":39,"tag":47,"props":1872,"children":1873},{},[1874],{"type":44,"value":1875},"Optional output directory specification:",{"type":39,"tag":187,"props":1877,"children":1879},{"className":1482,"code":1878,"language":1484,"meta":196,"style":196},"scripts\u002Fpackage_skill.py \u003Cpath\u002Fto\u002Fskill-folder> .\u002Fdist\n",[1880],{"type":39,"tag":194,"props":1881,"children":1882},{"__ignoreMap":196},[1883],{"type":39,"tag":653,"props":1884,"children":1885},{"class":655,"line":656},[1886,1890,1894,1898,1902,1906],{"type":39,"tag":653,"props":1887,"children":1888},{"style":666},[1889],{"type":44,"value":1852},{"type":39,"tag":653,"props":1891,"children":1892},{"style":660},[1893],{"type":44,"value":1501},{"type":39,"tag":653,"props":1895,"children":1896},{"style":788},[1897],{"type":44,"value":1861},{"type":39,"tag":653,"props":1899,"children":1900},{"style":708},[1901],{"type":44,"value":1866},{"type":39,"tag":653,"props":1903,"children":1904},{"style":660},[1905],{"type":44,"value":1516},{"type":39,"tag":653,"props":1907,"children":1908},{"style":788},[1909],{"type":44,"value":1910}," .\u002Fdist\n",{"type":39,"tag":47,"props":1912,"children":1913},{},[1914],{"type":44,"value":1915},"The packaging script will:",{"type":39,"tag":72,"props":1917,"children":1918},{},[1919,1952],{"type":39,"tag":76,"props":1920,"children":1921},{},[1922,1927,1929],{"type":39,"tag":117,"props":1923,"children":1924},{},[1925],{"type":44,"value":1926},"Validate",{"type":44,"value":1928}," the skill automatically, checking:",{"type":39,"tag":212,"props":1930,"children":1931},{},[1932,1937,1942,1947],{"type":39,"tag":76,"props":1933,"children":1934},{},[1935],{"type":44,"value":1936},"YAML frontmatter format and required fields",{"type":39,"tag":76,"props":1938,"children":1939},{},[1940],{"type":44,"value":1941},"Skill naming conventions and directory structure",{"type":39,"tag":76,"props":1943,"children":1944},{},[1945],{"type":44,"value":1946},"Description completeness and quality",{"type":39,"tag":76,"props":1948,"children":1949},{},[1950],{"type":44,"value":1951},"File organization and resource references",{"type":39,"tag":76,"props":1953,"children":1954},{},[1955,1960,1962,1968],{"type":39,"tag":117,"props":1956,"children":1957},{},[1958],{"type":44,"value":1959},"Package",{"type":44,"value":1961}," the skill if validation passes, creating a .skill file named after the skill (e.g., ",{"type":39,"tag":194,"props":1963,"children":1965},{"className":1964},[],[1966],{"type":44,"value":1967},"my-skill.skill",{"type":44,"value":1969},") that includes all files and maintains the proper directory structure for distribution. The .skill file is a zip file with a .skill extension.",{"type":39,"tag":47,"props":1971,"children":1972},{},[1973],{"type":44,"value":1974},"If validation fails, the script will report the errors and exit without creating a package. Fix any validation errors and run the packaging command again.",{"type":39,"tag":65,"props":1976,"children":1978},{"id":1977},"step-6-iterate",[1979],{"type":44,"value":1980},"Step 6: Iterate",{"type":39,"tag":47,"props":1982,"children":1983},{},[1984],{"type":44,"value":1985},"After testing the skill, users may request improvements. Often this happens right after using the skill, with fresh context of how the skill performed.",{"type":39,"tag":47,"props":1987,"children":1988},{},[1989],{"type":39,"tag":117,"props":1990,"children":1991},{},[1992],{"type":44,"value":1993},"Iteration workflow:",{"type":39,"tag":72,"props":1995,"children":1996},{},[1997,2002,2007,2012],{"type":39,"tag":76,"props":1998,"children":1999},{},[2000],{"type":44,"value":2001},"Use the skill on real tasks",{"type":39,"tag":76,"props":2003,"children":2004},{},[2005],{"type":44,"value":2006},"Notice struggles or inefficiencies",{"type":39,"tag":76,"props":2008,"children":2009},{},[2010],{"type":44,"value":2011},"Identify how SKILL.md or bundled resources should be updated",{"type":39,"tag":76,"props":2013,"children":2014},{},[2015],{"type":44,"value":2016},"Implement changes and test again",{"type":39,"tag":2018,"props":2019,"children":2020},"style",{},[2021],{"type":44,"value":2022},"html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"items":2024,"total":2126},[2025,2041,2057,2071,2088,2100,2114],{"slug":2026,"name":2026,"fn":2027,"description":2028,"org":2029,"tags":2030,"stars":22,"repoUrl":23,"updatedAt":2040},"codegen","generate code from UI specifications","Code generation utilities for json-render. Use when generating code from UI specs, building custom code exporters, traversing specs, or serializing props for @json-render\u002Fcodegen.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2031,2034,2037],{"name":2032,"slug":2033,"type":15},"Code Generation","code-generation",{"name":2035,"slug":2036,"type":15},"Engineering","engineering",{"name":2038,"slug":2039,"type":15},"JSON","json","2026-07-17T06:08:34.68038",{"slug":2042,"name":2042,"fn":2043,"description":2044,"org":2045,"tags":2046,"stars":22,"repoUrl":23,"updatedAt":2056},"devtools","inspect and debug generative UI components","Drop-in inspector panel for any json-render app. Use when the user wants to debug a generative UI, inspect the spec tree, edit state at runtime, see dispatched actions, follow stream patches live, browse a catalog, or pick DOM elements to find their spec keys. Triggers include \"add devtools\", \"debug json-render\", \"inspect the spec\", \"why is this element not rendering\", \"see the state at runtime\", or requests to tap streams \u002F capture action logs for `@json-render\u002Fdevtools`.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2047,2050,2053],{"name":2048,"slug":2049,"type":15},"Debugging","debugging",{"name":2051,"slug":2052,"type":15},"Frontend","frontend",{"name":2054,"slug":2055,"type":15},"UI Components","ui-components","2026-07-17T06:08:35.001228",{"slug":2058,"name":2058,"fn":2059,"description":2060,"org":2061,"tags":2062,"stars":22,"repoUrl":23,"updatedAt":2070},"directives","apply custom directives to JSON specs","Pre-built custom directives for json-render — formatting, math, string manipulation, and i18n. Use when working with @json-render\u002Fdirectives, defining custom directives with defineDirective, or adding $format, $math, $concat, $count, $truncate, $pluralize, $join, or $t to specs.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2063,2066,2067,2069],{"name":2064,"slug":2065,"type":15},"Automation","automation",{"name":2051,"slug":2052,"type":15},{"name":2068,"slug":2068,"type":15},"i18n",{"name":2038,"slug":2039,"type":15},"2026-07-17T06:04:05.866831",{"slug":2072,"name":2072,"fn":2073,"description":2074,"org":2075,"tags":2076,"stars":22,"repoUrl":23,"updatedAt":2087},"image","generate images from JSON specifications","Image renderer for json-render that turns JSON specs into SVG and PNG images via Satori. Use when working with @json-render\u002Fimage, generating OG images from JSON, creating social cards, or rendering AI-generated image specs.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2077,2080,2081,2084],{"name":2078,"slug":2079,"type":15},"Images","images",{"name":2038,"slug":2039,"type":15},{"name":2082,"slug":2083,"type":15},"Satori","satori",{"name":2085,"slug":2086,"type":15},"SVG","svg","2026-07-17T06:07:42.441875",{"slug":2089,"name":2089,"fn":2090,"description":2091,"org":2092,"tags":2093,"stars":22,"repoUrl":23,"updatedAt":2099},"ink","render JSON specs as terminal UIs","Ink terminal renderer for json-render that turns JSON specs into interactive terminal UIs. Use when working with @json-render\u002Fink, building terminal UIs from JSON, creating terminal component catalogs, or rendering AI-generated specs in the terminal.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2094,2097,2098],{"name":2095,"slug":2096,"type":15},"CLI","cli",{"name":2038,"slug":2039,"type":15},{"name":2054,"slug":2055,"type":15},"2026-07-17T06:08:28.678043",{"slug":2101,"name":2101,"fn":2102,"description":2103,"org":2104,"tags":2105,"stars":22,"repoUrl":23,"updatedAt":2113},"jotai","manage state with Jotai","Jotai adapter for json-render's StateStore interface. Use when integrating json-render with Jotai for state management via @json-render\u002Fjotai.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2106,2107,2110],{"name":2051,"slug":2052,"type":15},{"name":2108,"slug":2109,"type":15},"React","react",{"name":2111,"slug":2112,"type":15},"State Management","state-management","2026-07-17T06:05:48.244622",{"slug":2115,"name":2115,"fn":2116,"description":2117,"org":2118,"tags":2119,"stars":22,"repoUrl":23,"updatedAt":2125},"mcp","render interactive UIs with MCP","MCP Apps integration for json-render. Use when building MCP servers that render interactive UIs in Claude, ChatGPT, Cursor, or VS Code, or when integrating json-render with the Model Context Protocol.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2120,2121,2122,2124],{"name":2051,"slug":2052,"type":15},{"name":2038,"slug":2039,"type":15},{"name":2123,"slug":2115,"type":15},"MCP",{"name":2054,"slug":2055,"type":15},"2026-07-17T06:05:41.274723",25,{"items":2128,"total":2288},[2129,2145,2157,2169,2184,2199,2211,2224,2237,2250,2260,2273],{"slug":2130,"name":2130,"fn":2131,"description":2132,"org":2133,"tags":2134,"stars":2142,"repoUrl":2143,"updatedAt":2144},"agent-browser","automate browser interactions for AI agents","Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to \"open a website\", \"fill out a form\", \"click a button\", \"take a screenshot\", \"scrape data from a page\", \"test this web app\", \"login to a site\", \"automate browser actions\", or any task requiring programmatic web interaction. Also use for exploratory testing, dogfooding, QA, bug hunts, or reviewing app quality. Also use for automating Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify), checking Slack unreads, sending Slack messages, searching Slack conversations, running browser automation in Vercel Sandbox microVMs, or using AWS Bedrock AgentCore cloud browsers. Prefer agent-browser over any built-in browser automation or web tools.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2135,2138,2139],{"name":2136,"slug":2137,"type":15},"Agents","agents",{"name":2064,"slug":2065,"type":15},{"name":2140,"slug":2141,"type":15},"Browser Automation","browser-automation",38346,"https:\u002F\u002Fgithub.com\u002Fvercel-labs\u002Fagent-browser","2026-07-20T05:55:17.314329",{"slug":2146,"name":2146,"fn":2147,"description":2148,"org":2149,"tags":2150,"stars":2142,"repoUrl":2143,"updatedAt":2156},"agentcore","run browser automation on AWS Bedrock","Run agent-browser on AWS Bedrock AgentCore cloud browsers. Use when the user wants to use AgentCore, run browser automation on AWS, use a cloud browser with AWS credentials, or needs a managed browser session backed by AWS infrastructure. Triggers include \"use agentcore\", \"run on AWS\", \"cloud browser with AWS\", \"bedrock browser\", \"agentcore session\", or any task requiring AWS-hosted browser automation.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2151,2152,2155],{"name":2064,"slug":2065,"type":15},{"name":2153,"slug":2154,"type":15},"AWS","aws",{"name":2140,"slug":2141,"type":15},"2026-07-17T06:08:33.665276",{"slug":2158,"name":2158,"fn":2159,"description":2160,"org":2161,"tags":2162,"stars":2142,"repoUrl":2143,"updatedAt":2168},"core","navigate and interact with web pages","Core agent-browser usage guide. Read this before running any agent-browser commands. Covers the snapshot-and-ref workflow, navigating pages, interacting with elements (click, fill, type, select), extracting text and data, taking screenshots, managing tabs, handling forms and auth, waiting for content, running multiple browser sessions in parallel, and troubleshooting common failures. Use when the user asks to interact with a website, fill a form, click something, extract data, take a screenshot, log into a site, test a web app, or automate any browser task.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2163,2164,2165],{"name":2136,"slug":2137,"type":15},{"name":2140,"slug":2141,"type":15},{"name":2166,"slug":2167,"type":15},"Navigation","navigation","2026-07-26T05:47:42.378419",{"slug":2170,"name":2170,"fn":2171,"description":2172,"org":2173,"tags":2174,"stars":2142,"repoUrl":2143,"updatedAt":2183},"derive-client","reverse engineer internal APIs from browser traffic","Reverse-engineer a website's internal API by recording browser traffic into a HAR file, then generate a standalone client or CLI that calls the endpoints directly, with no browser needed after the first recording. Use when asked to \"derive a client\", \"build a CLI for \u003Csite>\", \"reverse engineer this site's API\", \"record network requests\", \"turn this site into an API\", or when the same site will be automated repeatedly and direct HTTP calls would beat driving the browser every time.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2175,2178,2179,2180],{"name":2176,"slug":2177,"type":15},"API Development","api-development",{"name":2064,"slug":2065,"type":15},{"name":2140,"slug":2141,"type":15},{"name":2181,"slug":2182,"type":15},"Web Scraping","web-scraping","2026-07-20T06:24:11.928835",{"slug":2185,"name":2185,"fn":2186,"description":2187,"org":2188,"tags":2189,"stars":2142,"repoUrl":2143,"updatedAt":2198},"dogfood","perform exploratory testing on web applications","Systematically explore and test a web application to find bugs, UX issues, and other problems. Use when asked to \"dogfood\", \"QA\", \"exploratory test\", \"find issues\", \"bug hunt\", \"test this app\u002Fsite\u002Fplatform\", or review the quality of a web application. Produces a structured report with full reproduction evidence -- step-by-step screenshots, repro videos, and detailed repro steps for every issue -- so findings can be handed directly to the responsible teams.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2190,2191,2192,2195],{"name":2140,"slug":2141,"type":15},{"name":2048,"slug":2049,"type":15},{"name":2193,"slug":2194,"type":15},"QA","qa",{"name":2196,"slug":2197,"type":15},"Testing","testing","2026-07-17T06:07:41.421482",{"slug":2200,"name":2200,"fn":2201,"description":2202,"org":2203,"tags":2204,"stars":2142,"repoUrl":2143,"updatedAt":2210},"electron","automate Electron desktop applications","Automate Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify, etc.) using agent-browser via Chrome DevTools Protocol. Use when the user needs to interact with an Electron app, automate a desktop app, connect to a running app, control a native app, or test an Electron application. Triggers include \"automate Slack app\", \"control VS Code\", \"interact with Discord app\", \"test this Electron app\", \"connect to desktop app\", or any task requiring automation of a native Electron application.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2205,2206,2207],{"name":2136,"slug":2137,"type":15},{"name":2140,"slug":2141,"type":15},{"name":2208,"slug":2209,"type":15},"Desktop","desktop","2026-07-17T06:08:28.007783",{"slug":2212,"name":2212,"fn":2213,"description":2214,"org":2215,"tags":2216,"stars":2142,"repoUrl":2143,"updatedAt":2223},"slack","interact with Slack workspaces","Interact with Slack workspaces using browser automation. Use when the user needs to check unread channels, navigate Slack, send messages, extract data, find information, search conversations, or automate any Slack task. Triggers include \"check my Slack\", \"what channels have unreads\", \"send a message to\", \"search Slack for\", \"extract from Slack\", \"find who said\", or any task requiring programmatic Slack interaction.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2217,2218,2221],{"name":2140,"slug":2141,"type":15},{"name":2219,"slug":2220,"type":15},"Messaging","messaging",{"name":2222,"slug":2212,"type":15},"Slack","2026-07-17T06:08:27.679015",{"slug":2225,"name":2225,"fn":2226,"description":2227,"org":2228,"tags":2229,"stars":2142,"repoUrl":2143,"updatedAt":2236},"vercel-sandbox","run browser automation in Vercel Sandbox","Run agent-browser + Chrome inside Vercel Sandbox microVMs for browser automation from any Vercel-deployed app. Use when the user needs browser automation in a Vercel app (Next.js, SvelteKit, Nuxt, Remix, Astro, etc.), wants to run headless Chrome without binary size limits, needs persistent browser sessions across commands, or wants ephemeral isolated browser environments. Triggers include \"Vercel Sandbox browser\", \"microVM Chrome\", \"agent-browser in sandbox\", \"browser automation on Vercel\", or any task requiring Chrome in a Vercel Sandbox.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2230,2231,2232,2233],{"name":2064,"slug":2065,"type":15},{"name":2140,"slug":2141,"type":15},{"name":2196,"slug":2197,"type":15},{"name":2234,"slug":2235,"type":15},"Vercel","vercel","2026-07-17T06:08:28.349899",{"slug":2238,"name":2238,"fn":2239,"description":2240,"org":2241,"tags":2242,"stars":2247,"repoUrl":2248,"updatedAt":2249},"deploy-to-vercel","deploy applications to Vercel","Deploy applications and websites to Vercel. Use when the user requests deployment actions like \"deploy my app\", \"deploy and give me the link\", \"push this live\", or \"create a preview deployment\".",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2243,2246],{"name":2244,"slug":2245,"type":15},"Deployment","deployment",{"name":2234,"slug":2235,"type":15},28993,"https:\u002F\u002Fgithub.com\u002Fvercel-labs\u002Fagent-skills","2026-07-17T06:08:41.18374",{"slug":2251,"name":2251,"fn":2252,"description":2253,"org":2254,"tags":2255,"stars":2247,"repoUrl":2248,"updatedAt":2259},"vercel-cli-with-tokens","manage Vercel projects via CLI","Deploy and manage projects on Vercel using token-based authentication. Use when working with Vercel CLI using access tokens rather than interactive login — e.g. \"deploy to vercel\", \"set up vercel\", \"add environment variables to vercel\".",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2256,2257,2258],{"name":2095,"slug":2096,"type":15},{"name":2244,"slug":2245,"type":15},{"name":2234,"slug":2235,"type":15},"2026-07-17T06:08:41.84179",{"slug":2261,"name":2261,"fn":2262,"description":2263,"org":2264,"tags":2265,"stars":2247,"repoUrl":2248,"updatedAt":2272},"vercel-composition-patterns","implement scalable React composition patterns","React composition patterns that scale. Use when refactoring components with boolean prop proliferation, building flexible component libraries, or designing reusable APIs. Triggers on tasks involving compound components, render props, context providers, or component architecture. Includes React 19 API changes.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2266,2269,2270,2271],{"name":2267,"slug":2268,"type":15},"Best Practices","best-practices",{"name":2051,"slug":2052,"type":15},{"name":2108,"slug":2109,"type":15},{"name":2054,"slug":2055,"type":15},"2026-07-17T06:05:40.576913",{"slug":2274,"name":2274,"fn":2275,"description":2276,"org":2277,"tags":2278,"stars":2247,"repoUrl":2248,"updatedAt":2287},"vercel-optimize","optimize Vercel project performance and costs","Use for Vercel cost and performance optimization on deployed projects, especially Next.js, SvelteKit, Nuxt, and limited Astro apps. Collect Vercel metrics, usage, project config, and code scan results first; investigate only metric-backed candidates; produce ranked recommendations grounded in verified files and version-aware Vercel\u002Fframework docs. Trigger for Vercel bill reduction, slow or expensive routes, caching opportunities, Function Invocations, Build Minutes, Fast Data Transfer, Core Web Vitals, Bot Management, Fluid compute, or cost breakdown requests.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[2279,2282,2283,2286],{"name":2280,"slug":2281,"type":15},"Cost Optimization","cost-optimization",{"name":2244,"slug":2245,"type":15},{"name":2284,"slug":2285,"type":15},"Performance","performance",{"name":2234,"slug":2235,"type":15},"2026-07-17T06:04:08.327515",100]