[{"data":1,"prerenderedAt":4997},["ShallowReactive",2],{"guide-en-skills\u002Fclaude-superpowers":3,"guide-siblings-en-skills":471,"guide-alt-en-skills\u002Fclaude-superpowers":4996},{"id":4,"title":5,"author":6,"body":7,"date":439,"description":440,"extension":441,"faq":442,"meta":461,"navigation":462,"order":463,"path":464,"readTime":465,"seo":466,"stem":467,"topic":468,"translationId":469,"updated":439,"__hash__":470},"guides\u002Fguides\u002Fskills\u002Fclaude-superpowers.md","Claude Code Superpowers: setup and how it works","Walma Engineering",{"type":8,"value":9,"toc":430},"minimark",[10,14,28,38,43,46,72,75,103,106,113,117,120,171,174,178,334,338,341,371,374,378,404,407,411,414,426],[11,12,13],"p",{},"Superpowers is the most widely used skill collection for Claude Code, with close to 300,000 GitHub stars in October 2026. It is not one skill but a working method: before Claude writes code it asks what you are trying to do, writes a short design, breaks the work into small tasks, builds them test-first and reviews the result. The skills trigger on their own, so after installing it you mostly notice that Claude stopped jumping straight into code.",[11,15,16,17,22,23,27],{},"This guide covers installation, the workflow step by step, what each skill does, what it costs and when to leave it off. If skills are new to you, start with ",[18,19,21],"a",{"href":20},"\u002Fen\u002Fguides\u002Fskills","what Claude skills are","; for other collections worth installing, see ",[18,24,26],{"href":25},"\u002Fen\u002Fguides\u002Fskills\u002Fbest-claude-skills","the best Claude skills",".",[11,29,30,31,37],{},"Facts below come from the ",[18,32,36],{"href":33,"rel":34},"https:\u002F\u002Fgithub.com\u002Fobra\u002Fsuperpowers",[35],"nofollow","obra\u002Fsuperpowers"," repository, checked on 12 October 2026.",[39,40,42],"h2",{"id":41},"install-superpowers-in-claude-code","Install Superpowers in Claude Code",[11,44,45],{},"Superpowers is in Anthropic's official plugin marketplace:",[47,48,53],"pre",{"className":49,"code":50,"language":51,"meta":52,"style":52},"language-bash shiki shiki-themes github-dark","\u002Fplugin install superpowers@claude-plugins-official\n","bash","",[54,55,56],"code",{"__ignoreMap":52},[57,58,61,65,69],"span",{"class":59,"line":60},"line",1,[57,62,64],{"class":63},"svObZ","\u002Fplugin",[57,66,68],{"class":67},"sU2Wk"," install",[57,70,71],{"class":67}," superpowers@claude-plugins-official\n",[11,73,74],{},"The author's own marketplace carries it too, along with related plugins:",[47,76,78],{"className":49,"code":77,"language":51,"meta":52,"style":52},"\u002Fplugin marketplace add obra\u002Fsuperpowers-marketplace\n\u002Fplugin install superpowers@superpowers-marketplace\n",[54,79,80,93],{"__ignoreMap":52},[57,81,82,84,87,90],{"class":59,"line":60},[57,83,64],{"class":63},[57,85,86],{"class":67}," marketplace",[57,88,89],{"class":67}," add",[57,91,92],{"class":67}," obra\u002Fsuperpowers-marketplace\n",[57,94,96,98,100],{"class":59,"line":95},2,[57,97,64],{"class":63},[57,99,68],{"class":67},[57,101,102],{"class":67}," superpowers@superpowers-marketplace\n",[11,104,105],{},"To check it is active, start describing something you want built: instead of writing code, Claude should start asking questions. That is the brainstorming skill.",[11,107,108,109,112],{},"Superpowers also installs in Codex, Cursor (",[54,110,111],{},"\u002Fadd-plugin superpowers","), Gemini CLI, GitHub Copilot CLI, OpenCode, Devin CLI and others. Each tool needs its own install.",[39,114,116],{"id":115},"the-workflow-step-by-step","The workflow, step by step",[11,118,119],{},"The core of Superpowers is seven skills that hand over to each other:",[121,122,123,131,137,143,153,159,165],"ol",{},[124,125,126,130],"li",{},[127,128,129],"strong",{},"brainstorming"," starts before anything is built. Claude asks what you want and why, then plays its understanding back so you can correct it. Small changes are agreed in chat; larger ones get a written design document.",[124,132,133,136],{},[127,134,135],{},"using-git-worktrees"," starts once the design is approved. It creates an isolated workspace on a new branch, runs the project setup and checks that the tests pass before anything changes.",[124,138,139,142],{},[127,140,141],{},"writing-plans"," turns the design into tasks of two to five minutes each, with exact file paths, the code to write and how to verify it.",[124,144,145,148,149,152],{},[127,146,147],{},"subagent-driven-development"," or ",[127,150,151],{},"executing-plans"," carries out the plan. The first gives each task to a fresh subagent and reviews after every task (most thorough). The second runs the whole plan in the current session with one review at the end (cheaper).",[124,154,155,158],{},[127,156,157],{},"test-driven-development"," applies during the work: write a failing test, watch it fail, write the minimum code, watch it pass, commit. Code written before its test gets deleted.",[124,160,161,164],{},[127,162,163],{},"requesting-code-review"," runs between tasks, checks the work against the plan and blocks progress on critical issues.",[124,166,167,170],{},[127,168,169],{},"finishing-a-development-branch"," verifies the tests and offers to merge, open a pull request, keep the branch or throw it away, then cleans up the worktree.",[11,172,173],{},"The repository describes these as \"mandatory workflows, not suggestions\": the agent checks for a relevant skill before every task. That is the point of it, and also the main reason some people turn it off.",[39,175,177],{"id":176},"every-skill-in-the-collection","Every skill in the collection",[179,180,181,194],"table",{},[182,183,184],"thead",{},[185,186,187,191],"tr",{},[188,189,190],"th",{},"Skill",[188,192,193],{},"What it does",[195,196,197,207,216,225,234,244,253,263,273,286,295,304,314,324],"tbody",{},[185,198,199,204],{},[200,201,202],"td",{},[127,203,129],{},[200,205,206],{},"Draws out intent before building",[185,208,209,213],{},[200,210,211],{},[127,212,141],{},[200,214,215],{},"Detailed, small-step implementation plans",[185,217,218,222],{},[200,219,220],{},[127,221,151],{},[200,223,224],{},"Runs a plan in one context with one final review",[185,226,227,231],{},[200,228,229],{},[127,230,147],{},[200,232,233],{},"One subagent per task, reviewed for spec compliance then code quality",[185,235,236,241],{},[200,237,238],{},[127,239,240],{},"dispatching-parallel-agents",[200,242,243],{},"Runs independent tasks with subagents in parallel",[185,245,246,250],{},[200,247,248],{},[127,249,157],{},[200,251,252],{},"Red, green, refactor, with a reference of testing anti-patterns",[185,254,255,260],{},[200,256,257],{},[127,258,259],{},"systematic-debugging",[200,261,262],{},"A four-phase root-cause process",[185,264,265,270],{},[200,266,267],{},[127,268,269],{},"verification-before-completion",[200,271,272],{},"Checks a fix actually works before calling it done",[185,274,275,283],{},[200,276,277,279,280],{},[127,278,163],{}," \u002F ",[127,281,282],{},"receiving-code-review",[200,284,285],{},"Review before merge, and how to act on feedback",[185,287,288,292],{},[200,289,290],{},[127,291,135],{},[200,293,294],{},"Parallel branches in separate working folders",[185,296,297,301],{},[200,298,299],{},[127,300,169],{},[200,302,303],{},"Merge, pull request or discard",[185,305,306,311],{},[200,307,308],{},[127,309,310],{},"diagnosing-superpowers",[200,312,313],{},"Reads a session transcript and explains what went wrong",[185,315,316,321],{},[200,317,318],{},[127,319,320],{},"writing-skills",[200,322,323],{},"Writes new skills the same way, with testing",[185,325,326,331],{},[200,327,328],{},[127,329,330],{},"using-superpowers",[200,332,333],{},"Introduces the system to the agent at the start of a session",[39,335,337],{"id":336},"what-it-costs","What it costs",[11,339,340],{},"Superpowers trades tokens for fewer wrong turns. A design, a plan, failing tests first and a review per task all cost tokens that \"just fix it\" does not.",[342,343,344,350,356],"ul",{},[124,345,346,349],{},[127,347,348],{},"Subagent-driven development is the expensive mode."," Each task gets a fresh subagent that has to read its way into the code, plus a review. In return, the main session's context stays clean, which is what lets it run for hours.",[124,351,352,355],{},[127,353,354],{},"executing-plans is the cheaper mode."," One session, one review at the end. Good for plans of a handful of tasks.",[124,357,358,361,362,366,367,27],{},[127,359,360],{},"On a Pro plan you will hit limits sooner."," If you pay per token, our ",[18,363,365],{"href":364},"\u002Fen\u002Fguides\u002Fclaude-code\u002Fclaude-api-pricing","Claude API pricing guide"," shows what a long agent session costs; on subscriptions, see ",[18,368,370],{"href":369},"\u002Fen\u002Fguides\u002Fclaude-code\u002Fclaude-code-pricing","Claude Code pricing",[11,372,373],{},"The author's README says it is \"not uncommon\" for the agent to work autonomously for a couple of hours from an approved plan. Budget for that before you let it.",[39,375,377],{"id":376},"when-superpowers-is-the-wrong-tool","When Superpowers is the wrong tool",[342,379,380,386,392,398],{},[124,381,382,385],{},[127,383,384],{},"One-line fixes and quick questions."," The process is overhead for a typo. Say it is a small change, or disable the plugin for that session.",[124,387,388,391],{},[127,389,390],{},"Exploratory spikes."," Test-first work slows you down when you do not yet know what you are building. Brainstorm, then switch it off while you prototype.",[124,393,394,397],{},[127,395,396],{},"Another methodology already installed."," Two collections with different rules for when to plan and when to test give Claude conflicting instructions. Pick one.",[124,399,400,403],{},[127,401,402],{},"Repositories without a test setup."," test-driven-development assumes you can run tests. Set that up first, or the workflow stalls at step 2.",[11,405,406],{},"When a session goes wrong (a skill fires when it should not, or the agent ignores its plan), ask Claude to \"figure out what went wrong with superpowers in this session\". The diagnosing-superpowers skill reads the transcript and points to the cause.",[39,408,410],{"id":409},"superpowers-across-a-team","Superpowers across a team",[11,412,413],{},"Superpowers pays off most when everyone uses it. Then every developer's agent plans, tests and reviews the same way, and pull requests from different people look alike.",[11,415,416,417,421,422,27],{},"That raises the usual team questions: which version is approved, who updates it, and whether the agent may merge on its own. Pin the plugin version, review updates before rolling them out, and decide whether finishing-a-development-branch may merge or only open pull requests. Our skills guide covers ",[18,418,420],{"href":419},"\u002Fen\u002Fguides\u002Fskills#rolling-skills-out-to-a-team","rolling skills out to a team"," in more detail, and Walma AI Hub distributes one approved set of skills and MCP servers to every developer's Claude client, with one policy and one log, hosted in the EU. ",[18,423,425],{"href":424},"\u002Fen\u002Fai-hub","See how it works",[427,428,429],"style",{},"html pre.shiki code .svObZ, html code.shiki .svObZ{--shiki-default:#B392F0}html pre.shiki code .sU2Wk, html code.shiki .sU2Wk{--shiki-default:#9ECBFF}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);}",{"title":52,"searchDepth":95,"depth":431,"links":432},3,[433,434,435,436,437,438],{"id":41,"depth":95,"text":42},{"id":115,"depth":95,"text":116},{"id":176,"depth":95,"text":177},{"id":336,"depth":95,"text":337},{"id":376,"depth":95,"text":377},{"id":409,"depth":95,"text":410},"2026-10-12","Superpowers is the most-used skill collection for Claude Code: brainstorm, plan, test first, then let subagents build. How to install it, the seven-step workflow, every skill in it, what it costs in tokens, and when it is the wrong tool.","md",[443,446,449,452,455,458],{"q":444,"a":445},"What is Superpowers for Claude Code?","Superpowers (github.com\u002Fobra\u002Fsuperpowers) is an open-source collection of agent skills that turns Claude Code into a fixed development process: it asks what you want before writing code, writes a design and a step-by-step plan, works test-first, reviews its own work between tasks, and finishes by verifying tests and offering to merge or open a pull request. It is MIT-licensed and built by Jesse Vincent and Prime Radiant.",{"q":447,"a":448},"How do I install Superpowers in Claude Code?","Run \u002Fplugin install superpowers@claude-plugins-official in Claude Code. It is in Anthropic's official plugin marketplace, so no extra marketplace is needed. The author's own marketplace (\u002Fplugin marketplace add obra\u002Fsuperpowers-marketplace) carries it too, alongside related plugins.",{"q":450,"a":451},"Does Superpowers work outside Claude Code?","Yes. The repository has install instructions for Codex (app and CLI), Cursor, Gemini CLI, GitHub Copilot CLI, OpenCode, Devin CLI, Factory Droid and several others. Install it separately in each tool you use.",{"q":453,"a":454},"Does Superpowers use more tokens?","Usually, yes. Brainstorming, written plans, test-first work and a review after each task all cost tokens a quick prompt would not spend. Subagent-driven development, where a fresh subagent handles each task, is the most thorough and the most expensive mode; executing-plans runs the plan in one session with a single review at the end and costs less.",{"q":456,"a":457},"How do I turn Superpowers off for a quick change?","Tell the agent it is a small change and you do not need a design or plan; small changes are agreed in chat rather than through a design document. For a session where you want none of it, disable the plugin with \u002Fplugin and re-enable it afterwards.",{"q":459,"a":460},"Does Superpowers send telemetry?","Only from one optional feature: the brainstorming skill's visual companion loads a logo from the author's website, which reveals the Superpowers version in use and nothing about your project. Set SUPERPOWERS_DISABLE_TELEMETRY, or Claude Code's DISABLE_TELEMETRY or CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC, to switch it off.",{},true,6,"\u002Fguides\u002Fskills\u002Fclaude-superpowers","9 min read",{"title":5,"description":440},"guides\u002Fskills\u002Fclaude-superpowers","skills",null,"89Vg-hqv78gUftnw95EDPzpsMk9XRhZQoKIsDNMFRc0",[472,1546,1862,2530,3753,4191,4693],{"id":473,"title":474,"author":6,"body":475,"date":1511,"description":1512,"extension":441,"faq":1513,"meta":1538,"navigation":462,"order":1539,"path":1540,"readTime":1541,"seo":1542,"stem":1543,"topic":468,"translationId":1544,"updated":439,"__hash__":1545},"guides\u002Fguides\u002Fskills\u002Findex.md","Claude skills: how to install, use and write them",{"type":8,"value":476,"toc":1494},[477,484,491,495,498,504,508,511,534,537,541,544,552,555,699,706,709,715,728,732,738,758,765,769,772,867,870,873,904,909,919,1010,1013,1067,1071,1148,1155,1159,1162,1168,1174,1180,1186,1195,1201,1205,1214,1232,1243,1246,1252,1265,1271,1289,1292,1296,1299,1353,1357,1360,1401,1405,1421,1424,1428,1440,1446,1452,1460,1464,1467,1473,1479,1485,1491],[11,478,479,480,483],{},"A skill is a folder. Inside it is a Markdown file called ",[54,481,482],{},"SKILL.md"," that tells Claude how to do a specific job, and optionally some scripts and reference documents that help. When a task matches the skill's description, Claude reads the file and follows it. When it does not, the skill costs nothing.",[11,485,486,487,27],{},"That is the whole mechanism, and it turns out to be the most practical way anyone has found to give an AI agent expertise that survives from one session to the next. This guide covers how skills work, how to install them in Claude Code, Claude.ai and the API, how to write one that actually improves results, and what to consider before a team adopts them. For how skills relate to MCP, see ",[18,488,490],{"href":489},"\u002Fen\u002Fguides\u002Fskills\u002Fclaude-skills-vs-mcp","Claude skills vs MCP",[39,492,494],{"id":493},"why-skills-exist","Why skills exist",[11,496,497],{},"Prompts do not scale. A good prompt for \"review this pull request the way our team does it\" is two pages long. Nobody pastes two pages into every session, so the knowledge lives in someone's head, or in a wiki the agent never reads.",[11,499,500,501,503],{},"Skills package that knowledge once, next to the code, and load it only when needed. Anthropic launched them in October 2025 for Claude.ai, Claude Code and the API. In December 2025 the format was published as an open specification, Agent Skills, and other coding agents adopted it. In practice, a ",[54,502,482],{}," you write for Claude Code today also works in Codex, Cursor and Copilot with minor differences.",[39,505,507],{"id":506},"how-a-skill-works","How a skill works",[11,509,510],{},"Three things happen at different times, which is why skills scale.",[121,512,513,519,528],{},[124,514,515,518],{},[127,516,517],{},"At startup",", the agent reads only the frontmatter of every skill it knows about: the name and a one-line description. For fifty skills that is a few hundred tokens.",[124,520,521,524,525,527],{},[127,522,523],{},"When a task matches a description",", the agent reads the full ",[54,526,482],{}," into context. That is when the instructions start to matter.",[124,529,530,533],{},[127,531,532],{},"When the instructions say so",", the agent opens reference files or runs scripts in the skill folder. A skill for generating Excel files might carry a Python script that does the actual writing, so the model does not have to reinvent it.",[11,535,536],{},"This is called progressive disclosure. It is the reason you can have many skills installed without paying for all of them every turn.",[39,538,540],{"id":539},"anatomy-of-a-skill","Anatomy of a skill",[11,542,543],{},"The minimum is one file:",[47,545,550],{"className":546,"code":548,"language":549},[547],"language-text","release-notes\u002F\n└── SKILL.md\n","text",[54,551,548],{"__ignoreMap":52},[11,553,554],{},"With this content:",[47,556,560],{"className":557,"code":558,"language":559,"meta":52,"style":52},"language-markdown shiki shiki-themes github-dark","---\nname: release-notes\ndescription: Write release notes from merged pull requests in our house style. Use when asked to draft, write or prepare release notes or a changelog for a version.\n---\n\n# Release notes\n\n## When to use\nThe user asks for release notes, a changelog or \"what shipped\" for a version or date range.\n\n## Steps\n1. List merged PRs for the range using the GitHub tools. Ignore PRs labelled `chore` or `internal`.\n2. Group by: New, Improved, Fixed. One line each, present tense, no PR numbers in the line.\n3. Lead with the change that affects the most users.\n4. End with an \"Upgrade notes\" section only if any PR is labelled `breaking`.\n\n## Style\nShort sentences. No exclamation marks. British spelling. Product names as in `docs\u002Fstyle.md`.\n","markdown",[54,561,562,568,574,579,584,590,595,600,606,612,617,623,645,654,663,677,682,688],{"__ignoreMap":52},[57,563,564],{"class":59,"line":60},[57,565,567],{"class":566},"sIZOC","---\n",[57,569,570],{"class":59,"line":95},[57,571,573],{"class":572},"s95oV","name: release-notes\n",[57,575,576],{"class":59,"line":431},[57,577,578],{"class":572},"description: Write release notes from merged pull requests in our house style. Use when asked to draft, write or prepare release notes or a changelog for a version.\n",[57,580,582],{"class":59,"line":581},4,[57,583,567],{"class":566},[57,585,587],{"class":59,"line":586},5,[57,588,589],{"emptyLinePlaceholder":462},"\n",[57,591,592],{"class":59,"line":463},[57,593,594],{"class":566},"# Release notes\n",[57,596,598],{"class":59,"line":597},7,[57,599,589],{"emptyLinePlaceholder":462},[57,601,603],{"class":59,"line":602},8,[57,604,605],{"class":566},"## When to use\n",[57,607,609],{"class":59,"line":608},9,[57,610,611],{"class":572},"The user asks for release notes, a changelog or \"what shipped\" for a version or date range.\n",[57,613,615],{"class":59,"line":614},10,[57,616,589],{"emptyLinePlaceholder":462},[57,618,620],{"class":59,"line":619},11,[57,621,622],{"class":566},"## Steps\n",[57,624,626,630,633,637,639,642],{"class":59,"line":625},12,[57,627,629],{"class":628},"s9osk","1.",[57,631,632],{"class":572}," List merged PRs for the range using the GitHub tools. Ignore PRs labelled ",[57,634,636],{"class":635},"sDLfK","`chore`",[57,638,148],{"class":572},[57,640,641],{"class":635},"`internal`",[57,643,644],{"class":572},".\n",[57,646,648,651],{"class":59,"line":647},13,[57,649,650],{"class":628},"2.",[57,652,653],{"class":572}," Group by: New, Improved, Fixed. One line each, present tense, no PR numbers in the line.\n",[57,655,657,660],{"class":59,"line":656},14,[57,658,659],{"class":628},"3.",[57,661,662],{"class":572}," Lead with the change that affects the most users.\n",[57,664,666,669,672,675],{"class":59,"line":665},15,[57,667,668],{"class":628},"4.",[57,670,671],{"class":572}," End with an \"Upgrade notes\" section only if any PR is labelled ",[57,673,674],{"class":635},"`breaking`",[57,676,644],{"class":572},[57,678,680],{"class":59,"line":679},16,[57,681,589],{"emptyLinePlaceholder":462},[57,683,685],{"class":59,"line":684},17,[57,686,687],{"class":566},"## Style\n",[57,689,691,694,697],{"class":59,"line":690},18,[57,692,693],{"class":572},"Short sentences. No exclamation marks. British spelling. Product names as in ",[57,695,696],{"class":635},"`docs\u002Fstyle.md`",[57,698,644],{"class":572},[11,700,701,702,705],{},"The ",[54,703,704],{},"description"," is doing real work. It is what the agent matches against, so it should say both what the skill does and when to use it, in the words a user would actually type.",[11,707,708],{},"A richer skill adds files:",[47,710,713],{"className":711,"code":712,"language":549},[547],"release-notes\u002F\n├── SKILL.md\n├── examples\u002F\n│   └── v2.4.md\n└── scripts\u002F\n    └── list_prs.sh\n",[54,714,712],{"__ignoreMap":52},[11,716,717,719,720,723,724,727],{},[54,718,482],{}," refers to them (\"see ",[54,721,722],{},"examples\u002Fv2.4.md"," for the expected format\", \"run ",[54,725,726],{},"scripts\u002Flist_prs.sh \u003Cfrom> \u003Cto>","\"). The agent reads or executes them only when it gets to that step.",[39,729,731],{"id":730},"is-a-skill-just-a-prompt","Is a skill just a prompt?",[11,733,734,735,737],{},"It is the most common question about skills, and the honest answer is: the core of it is. ",[54,736,482],{}," is instructions written in Markdown, and a model reads it the way it reads any prompt. Three things make it more than a prompt in practice:",[342,739,740,746,752],{},[124,741,742,745],{},[127,743,744],{},"It loads itself."," You do not paste it. The agent reads it only when a request matches its description, so fifty skills cost almost nothing until one is needed.",[124,747,748,751],{},[127,749,750],{},"It carries tools."," Scripts, templates and reference files sit next to the instructions, and the agent runs or opens them when a step calls for it. A prompt cannot ship a Python script that parses your invoice format correctly every time.",[124,753,754,757],{},[127,755,756],{},"It is a file in a repository."," You can review it in a pull request, version it, roll it back and share it with a team. That is what turns one person's good prompt into a team convention.",[11,759,760,761,764],{},"If a skill would be one paragraph with no files, a line in ",[54,762,763],{},"CLAUDE.md"," or a slash command may be the better tool. The comparison is further down.",[39,766,768],{"id":767},"how-to-install-skills-in-claude-code","How to install skills in Claude Code",[11,770,771],{},"Claude Code looks for skills in several places. These are the ones you will use:",[179,773,774,787],{},[182,775,776],{},[185,777,778,781,784],{},[188,779,780],{},"Where",[188,782,783],{},"Path",[188,785,786],{},"Who gets it",[195,788,789,805,820,835,854],{},[185,790,791,796,802],{},[200,792,793],{},[127,794,795],{},"Project",[200,797,798,801],{},[54,799,800],{},".claude\u002Fskills\u002F\u003Cname>\u002FSKILL.md"," in the repository",[200,803,804],{},"Everyone who clones the repo. This is where team conventions belong.",[185,806,807,812,817],{},[200,808,809],{},[127,810,811],{},"Personal",[200,813,814],{},[54,815,816],{},"~\u002F.claude\u002Fskills\u002F\u003Cname>\u002FSKILL.md",[200,818,819],{},"You, in every project on this machine.",[185,821,822,827,832],{},[200,823,824],{},[127,825,826],{},"Nested",[200,828,829],{},[54,830,831],{},"\u003Csubdir>\u002F.claude\u002Fskills\u002F\u003Cname>\u002FSKILL.md",[200,833,834],{},"Loaded when Claude works on files in that part of a monorepo.",[185,836,837,842,847],{},[200,838,839],{},[127,840,841],{},"Plugin",[200,843,844,845],{},"Installed from a marketplace with ",[54,846,64],{},[200,848,849,850,853],{},"Named ",[54,851,852],{},"\u002Fplugin-name:skill-name",". Plugins bundle skills with commands, hooks and MCP servers.",[185,855,856,861,864],{},[200,857,858],{},[127,859,860],{},"Enterprise",[200,862,863],{},"Managed settings, pushed by an admin",[200,865,866],{},"Everyone in the organisation.",[11,868,869],{},"When two skills share a name, enterprise wins over personal, and personal wins over project. That is useful for an organisation that wants one approved version of a skill to override local copies.",[11,871,872],{},"Once a skill is in place there is nothing to enable. Three ways to use it:",[342,874,875,881,894],{},[124,876,877,880],{},[127,878,879],{},"Automatically."," Type a request that matches its description and Claude loads it.",[124,882,883,886,887,890,891,27],{},[127,884,885],{},"Explicitly."," Type ",[54,888,889],{},"\u002Fskill-name"," at the start of a message, with arguments if the skill takes them: ",[54,892,893],{},"\u002Frelease-notes v2.4",[124,895,896,899,900,903],{},[127,897,898],{},"List what is loaded"," with ",[54,901,902],{},"\u002Fskills",", which shows every available skill and its description.",[905,906,908],"h3",{"id":907},"the-frontmatter-fields-worth-knowing","The frontmatter fields worth knowing",[11,910,911,912,915,916,918],{},"Only ",[54,913,914],{},"name"," and ",[54,917,704],{}," are needed to start, but a few optional fields in Claude Code change how a skill behaves:",[179,920,921,930],{},[182,922,923],{},[185,924,925,928],{},[188,926,927],{},"Field",[188,929,193],{},[195,931,932,945,959,973,987,1000],{},[185,933,934,938],{},[200,935,936],{},[54,937,704],{},[200,939,940,941,944],{},"What the skill does and when to use it. Claude matches requests against it. Together with ",[54,942,943],{},"when_to_use"," it is truncated at 1,536 characters in the skill listing.",[185,946,947,952],{},[200,948,949],{},[54,950,951],{},"disable-model-invocation: true",[200,953,954,955,958],{},"Only you can trigger it with ",[54,956,957],{},"\u002Fname","; Claude never runs it on its own. Use it for anything with side effects, such as deploys.",[185,960,961,966],{},[200,962,963],{},[54,964,965],{},"user-invocable: false",[200,967,968,969,972],{},"The opposite: hidden from the ",[54,970,971],{},"\u002F"," menu, used only when Claude decides it is relevant.",[185,974,975,980],{},[200,976,977],{},[54,978,979],{},"allowed-tools",[200,981,982,983,986],{},"Tools the skill may use without asking, for example specific ",[54,984,985],{},"git"," commands.",[185,988,989,994],{},[200,990,991],{},[54,992,993],{},"paths",[200,995,996,997,27],{},"Glob patterns that limit when the skill auto-loads, for example only for files under ",[54,998,999],{},"infra\u002F",[185,1001,1002,1007],{},[200,1003,1004],{},[54,1005,1006],{},"context: fork",[200,1008,1009],{},"Runs the skill in a separate subagent with its own context, so a long procedure does not fill the main conversation.",[11,1011,1012],{},"A deploy skill, for example, is safer like this:",[47,1014,1018],{"className":1015,"code":1016,"language":1017,"meta":52,"style":52},"language-yaml shiki shiki-themes github-dark","---\nname: deploy\ndescription: Deploy the application to production after the release checklist passes.\ndisable-model-invocation: true\nallowed-tools: Bash(git status *) Bash(git tag *)\n---\n","yaml",[54,1019,1020,1024,1035,1044,1054,1063],{"__ignoreMap":52},[57,1021,1022],{"class":59,"line":60},[57,1023,567],{"class":63},[57,1025,1026,1029,1032],{"class":59,"line":95},[57,1027,914],{"class":1028},"s4JwU",[57,1030,1031],{"class":572},": ",[57,1033,1034],{"class":67},"deploy\n",[57,1036,1037,1039,1041],{"class":59,"line":431},[57,1038,704],{"class":1028},[57,1040,1031],{"class":572},[57,1042,1043],{"class":67},"Deploy the application to production after the release checklist passes.\n",[57,1045,1046,1049,1051],{"class":59,"line":581},[57,1047,1048],{"class":1028},"disable-model-invocation",[57,1050,1031],{"class":572},[57,1052,1053],{"class":635},"true\n",[57,1055,1056,1058,1060],{"class":59,"line":586},[57,1057,979],{"class":1028},[57,1059,1031],{"class":572},[57,1061,1062],{"class":67},"Bash(git status *) Bash(git tag *)\n",[57,1064,1065],{"class":59,"line":463},[57,1066,567],{"class":63},[39,1068,1070],{"id":1069},"skills-in-claudeai-the-api-and-managed-agents","Skills in Claude.ai, the API and Managed Agents",[179,1072,1073,1086],{},[182,1074,1075],{},[185,1076,1077,1080,1083],{},[188,1078,1079],{},"Surface",[188,1081,1082],{},"How skills get there",[188,1084,1085],{},"Notes",[195,1087,1088,1101,1115,1132],{},[185,1089,1090,1095,1098],{},[200,1091,1092],{},[127,1093,1094],{},"Claude Code",[200,1096,1097],{},"Folders on disk, plugins, managed settings",[200,1099,1100],{},"See above.",[185,1102,1103,1109,1112],{},[200,1104,1105,1108],{},[127,1106,1107],{},"Claude.ai"," (web, desktop)",[200,1110,1111],{},"Settings: enable Anthropic's built-in skills or upload your own as a zip",[200,1113,1114],{},"Available on Free, Pro, Max, Team and Enterprise. Requires code execution to be enabled. Team and Enterprise owners can provision skills for the whole organisation.",[185,1116,1117,1122,1129],{},[200,1118,1119],{},[127,1120,1121],{},"Claude API",[200,1123,1124,1125,1128],{},"Upload once through the Skills API (",[54,1126,1127],{},"\u002Fv1\u002Fskills","), reference by ID on a request that uses the code execution container",[200,1130,1131],{},"How a production agent gets the same expertise a developer has in Claude Code.",[185,1133,1134,1139,1142],{},[200,1135,1136],{},[127,1137,1138],{},"Claude Managed Agents",[200,1140,1141],{},"Part of the agent definition, alongside model, tools and MCP servers",[200,1143,1144,1145,27],{},"See ",[18,1146,1138],{"href":1147},"\u002Fen\u002Fguides\u002Fai-agents\u002Fclaude-managed-agents",[11,1149,1150,1151,1154],{},"The built-in skills in Claude.ai cover document formats: Word, Excel, PowerPoint and PDF. They are the reason Claude can produce a real ",[54,1152,1153],{},".xlsx"," with formulas instead of a table pasted into a chat.",[39,1156,1158],{"id":1157},"writing-a-skill-that-works","Writing a skill that works",[11,1160,1161],{},"Most skills we see in the wild are too long, too vague, or both. The ones that reliably change the agent's behaviour share a few properties.",[11,1163,1164,1167],{},[127,1165,1166],{},"The description is precise."," \"Helps with documents\" matches everything and nothing. \"Convert a DOCX contract into our redline format. Use when asked to redline, mark up or compare contract versions\" matches the right requests.",[11,1169,1170,1173],{},[127,1171,1172],{},"The body is procedural."," Numbered steps, the order they should happen in, and what \"done\" looks like. The agent is good at following a procedure and bad at inferring one from a description of values.",[11,1175,1176,1179],{},[127,1177,1178],{},"Decisions are explicit."," If the task has a fork (\"if the repo has a CHANGELOG.md, append to it; otherwise create one\"), write the fork. The agent will otherwise pick one at random and be confidently wrong half the time.",[11,1181,1182,1185],{},[127,1183,1184],{},"Deterministic work goes in scripts."," Anything the model would do by generating code (parse a file format, call an API with the right pagination, compute a checksum) is more reliable as a script the skill runs. The model then orchestrates rather than reinvents.",[11,1187,1188,1191,1192,1194],{},[127,1189,1190],{},"It is short enough to read."," Under 500 lines for ",[54,1193,482],{},". Anything longer moves into reference files the agent opens on demand.",[11,1196,1197,1200],{},[127,1198,1199],{},"It has an example."," One example of the expected output is worth a page of description.",[39,1202,1204],{"id":1203},"a-worked-example-ux-lint","A worked example: ux-lint",[11,1206,1207,1208,1213],{},"The properties above are easier to see in a skill that exists. We publish ",[18,1209,1212],{"href":1210,"rel":1211},"https:\u002F\u002Fgithub.com\u002FWalma-Labs\u002Fux-lint",[35],"ux-lint",", an open-source UX linter for coding agents, under the same Agent Skills format. It installs like any other skill:",[47,1215,1217],{"className":49,"code":1216,"language":51,"meta":52,"style":52},"npx skills add Walma-Labs\u002Fux-lint\n",[54,1218,1219],{"__ignoreMap":52},[57,1220,1221,1224,1227,1229],{"class":59,"line":60},[57,1222,1223],{"class":63},"npx",[57,1225,1226],{"class":67}," skills",[57,1228,89],{"class":67},[57,1230,1231],{"class":67}," Walma-Labs\u002Fux-lint\n",[11,1233,1234,1235,1239,1240,27],{},"Then you ask for it in the words you would use anyway: ",[1236,1237,1238],"em",{},"\"review the signup page for UX problems\"",", ",[1236,1241,1242],{},"\"ux check before I merge\"",[11,1244,1245],{},"It is worth reading as an example because of four choices, each one an instance of a rule from the previous section.",[11,1247,1248,1251],{},[127,1249,1250],{},"The deterministic work is a script, not a prompt."," The skill ships a Playwright script that opens the page in a real browser and measures it: tap-target sizes and spacing, elements that cannot actually be clicked, horizontal overflow, FCP and LCP, click-to-paint latency, layout shift. The model never estimates a pixel value. It receives numbers and reasons about them. That is the difference between a skill that works and a design skill that guesses from source code.",[11,1253,1254,1260,1261,1264],{},[127,1255,1256,1257,1259],{},"The reference material sits beside ",[54,1258,482],{},", not inside it."," Every UX principle is traced to its primary research in a ",[54,1262,1263],{},"references\u002F"," folder, which the agent opens when it needs to justify a finding. The instruction file stays short enough for the agent to actually follow.",[11,1266,1267,1270],{},[127,1268,1269],{},"The instructions constrain the model, not just guide it."," Each principle is labelled with the strength of its evidence, and the skill documents where popular UX advice overstates the research. It will not flag a navigation bar for \"violating Miller's Law\", because that law is about memory, not visible menu items. Writing the boundary into the skill is what stops a confident-sounding wrong answer.",[11,1272,1273,1276,1277,1280,1281,1284,1285,1288],{},[127,1274,1275],{},"One skill, three modes."," ",[54,1278,1279],{},"review"," produces a ranked report, ",[54,1282,1283],{},"check"," a pass\u002Ffail verdict for pre-merge, ",[54,1286,1287],{},"fix"," applies the changes and re-measures. The modes map to when people actually ask, which is the practical version of \"the description is precise\".",[11,1290,1291],{},"If you are writing your first skill, cloning the shape of a working one is faster than starting from the specification.",[39,1293,1295],{"id":1294},"when-a-skill-does-not-trigger","When a skill does not trigger",[11,1297,1298],{},"The most common complaint about skills is that Claude ignores them. Work through these in order:",[121,1300,1301,1312,1320,1326,1336,1347],{},[124,1302,1303,1306,1307,1309,1310,27],{},[127,1304,1305],{},"Is it loaded?"," Run ",[54,1308,902],{}," in Claude Code. If it is not listed, the folder is in the wrong place or the file is not called ",[54,1311,482],{},[124,1313,1314,886,1317,1319],{},[127,1315,1316],{},"Does it work when called directly?",[54,1318,889],{},". If that works, the skill is fine and the problem is matching.",[124,1321,1322,1325],{},[127,1323,1324],{},"Does the description use the user's words?"," Claude matches the request against the description. If people say \"changelog\" and the description says \"release documentation\", it will not fire. Name the task and the phrases people actually type.",[124,1327,1328,1331,1332,1335],{},[127,1329,1330],{},"Is the frontmatter valid YAML?"," A skill with malformed frontmatter loads with empty metadata, so it has no description to match. Running Claude Code with ",[54,1333,1334],{},"--debug"," shows parse errors.",[124,1337,1338,1276,1341,1343,1344,1346],{},[127,1339,1340],{},"Is automatic use switched off?",[54,1342,951],{}," means only ",[54,1345,889],{}," runs it.",[124,1348,1349,1352],{},[127,1350,1351],{},"It worked once and then stopped?"," Claude Code reads a skill into the conversation once and does not re-read the file on later turns. Phrase instructions so they apply for the whole task, or invoke the skill again.",[39,1354,1356],{"id":1355},"before-you-install-someone-elses-skill","Before you install someone else's skill",[11,1358,1359],{},"A third-party skill runs with your permissions. It can contain scripts, and its instructions can steer the agent. Install skills only from sources you trust, and review them first. A five-minute review catches most problems:",[342,1361,1362,1371,1381,1389,1395],{},[124,1363,1364,1370],{},[127,1365,1366,1367,1369],{},"Read ",[54,1368,482],{}," in full."," Look for instructions that send data anywhere (\"post the summary to…\", \"upload to…\"), touch credentials, or tell the agent to skip confirmations.",[124,1372,1373,1376,1377,1380],{},[127,1374,1375],{},"Read every script."," Network calls, ",[54,1378,1379],{},"curl | sh",", writes outside the project, and anything obfuscated are reasons to stop.",[124,1382,1383,1388],{},[127,1384,1385,1386,27],{},"Check ",[54,1387,979],{}," A skill that pre-approves broad shell access deserves more scrutiny than one that only reads files.",[124,1390,1391,1394],{},[127,1392,1393],{},"Pin a version."," Install from a specific commit or release, not a moving branch, so the skill you reviewed is the skill that runs.",[124,1396,1397,1400],{},[127,1398,1399],{},"Prefer a team copy."," Fork reviewed skills into your own repository or internal marketplace, and update them deliberately.",[39,1402,1404],{"id":1403},"where-to-find-skills","Where to find skills",[11,1406,1407,1408,1410,1411,1415,1416,1420],{},"Anthropic maintains a public repository of official skills on GitHub, covering document formats, presentations and a few developer workflows. Our shortlist of ",[18,1409,26],{"href":25}," covers the official ones and the community skills worth installing, ",[18,1412,1414],{"href":1413},"\u002Fen\u002Fguides\u002Fskills\u002Fclaude-marketing-skills","Claude marketing skills"," covers the marketing collections, and ",[18,1417,1419],{"href":1418},"\u002Fen\u002Fguides\u002Fskills\u002Fclaude-skill-creator","skill-creator"," is the way to build and test your own. Claude Code's plugin marketplaces are the main distribution channel for third-party skills. Community \"awesome\" lists collect more.",[11,1422,1423],{},"The rule is the same as for any code from the internet: read it before you install it. A skill can contain scripts, and scripts run. A skill's instructions can also tell the agent to do things you would not want (\"post the results to this webhook\"). For a team, that means an approved set rather than an open marketplace, which we come back to below.",[39,1425,1427],{"id":1426},"skills-vs-the-alternatives","Skills vs the alternatives",[11,1429,1430,1433,1434,1436,1437,1439],{},[127,1431,1432],{},"Skills vs CLAUDE.md."," A ",[54,1435,763],{}," file is always loaded and should hold what the agent must always know: how to build, test, and the non-negotiable conventions. Skills hold task-specific knowledge that would bloat that file. If it applies to every session, put it in ",[54,1438,763],{},"; if it applies to one kind of task, make it a skill.",[11,1441,1442,1445],{},[127,1443,1444],{},"Skills vs slash commands."," Commands are user-triggered shortcuts for a prompt. Skills are triggered by the agent when relevant and can carry files and scripts. Many teams start with commands and graduate to skills when the prompt grows or needs supporting material.",[11,1447,1448,1451],{},[127,1449,1450],{},"Skills vs subagents."," A subagent is a separate context with its own tools, used for isolation or parallelism. A skill is knowledge loaded into the current context. A subagent can use skills.",[11,1453,1454,1457,1458,27],{},[127,1455,1456],{},"Skills vs MCP."," MCP gives access to systems; skills give know-how. The full comparison is in ",[18,1459,490],{"href":489},[39,1461,1463],{"id":1462},"rolling-skills-out-to-a-team","Rolling skills out to a team",[11,1465,1466],{},"Three things change when skills go from one developer to a department.",[11,1468,1469,1472],{},[127,1470,1471],{},"Provenance."," Which skills are approved, who wrote them, what scripts they contain. A curated internal marketplace, or a repository of reviewed skills, replaces \"install whatever the blog said\".",[11,1474,1475,1478],{},[127,1476,1477],{},"Consistency."," The value of a skill for \"how we review PRs\" is that everyone's agent reviews PRs the same way. That only holds if the skill is the same everywhere and updated centrally.",[11,1480,1481,1484],{},[127,1482,1483],{},"Policy."," A skill that says \"deploy to production after the tests pass\" should only run for people allowed to deploy. Skills do not enforce that; the tools they call do. This is where skills meet the gateway: the skill describes the procedure, the gateway decides whether this user's agent may call the deploy tool.",[11,1486,1487,1488,27],{},"Walma AI Hub handles the last part. Approved skills and MCP servers are distributed through one signed client, policies decide who can call which tools, and every call is logged, inside the customer's own EU region. If you are standardising skills across a team, ",[18,1489,1490],{"href":424},"book a walkthrough",[427,1492,1493],{},"html pre.shiki code .sIZOC, html code.shiki .sIZOC{--shiki-default:#79B8FF;--shiki-default-font-weight:bold}html pre.shiki code .s95oV, html code.shiki .s95oV{--shiki-default:#E1E4E8}html pre.shiki code .s9osk, html code.shiki .s9osk{--shiki-default:#FFAB70}html pre.shiki code .sDLfK, html code.shiki .sDLfK{--shiki-default:#79B8FF}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 pre.shiki code .svObZ, html code.shiki .svObZ{--shiki-default:#B392F0}html pre.shiki code .sU2Wk, html code.shiki .sU2Wk{--shiki-default:#9ECBFF}html pre.shiki code .s4JwU, html code.shiki .s4JwU{--shiki-default:#85E89D}",{"title":52,"searchDepth":95,"depth":431,"links":1495},[1496,1497,1498,1499,1500,1503,1504,1505,1506,1507,1508,1509,1510],{"id":493,"depth":95,"text":494},{"id":506,"depth":95,"text":507},{"id":539,"depth":95,"text":540},{"id":730,"depth":95,"text":731},{"id":767,"depth":95,"text":768,"children":1501},[1502],{"id":907,"depth":431,"text":908},{"id":1069,"depth":95,"text":1070},{"id":1157,"depth":95,"text":1158},{"id":1203,"depth":95,"text":1204},{"id":1294,"depth":95,"text":1295},{"id":1355,"depth":95,"text":1356},{"id":1403,"depth":95,"text":1404},{"id":1426,"depth":95,"text":1427},{"id":1462,"depth":95,"text":1463},"2026-09-11","How to install a skill in Claude Code and Claude.ai, how skills work, how to write your own SKILL.md, why a skill does not trigger, and what to check before rolling skills out to a team.",[1514,1517,1520,1523,1526,1529,1532,1535],{"q":1515,"a":1516},"What are Claude skills?","Skills are folders containing a SKILL.md file with instructions, plus optional scripts and reference files, that teach Claude how to perform a specific task. Claude loads a skill only when it is relevant, so you can have many without filling the context window.",{"q":1518,"a":1519},"How do I install a skill in Claude Code?","Put the skill folder in .claude\u002Fskills\u002F inside your project (shared with the team via git) or in ~\u002F.claude\u002Fskills\u002F for personal use. Skills can also come from plugins installed from a marketplace. Claude Code picks them up automatically.",{"q":1521,"a":1522},"Are skills the same as MCP?","No. Skills are instructions and scripts that tell the model how to do something. MCP servers give the model access to external tools and data. A skill often uses MCP tools to do its work. They are complementary.",{"q":1524,"a":1525},"Do skills work with ChatGPT, Codex or Cursor?","The SKILL.md format was published as an open specification called Agent Skills in late 2025, and Codex, Cursor, GitHub Copilot and Gemini CLI have adopted it. A well-written skill is portable across those tools, with some differences in how they are discovered.",{"q":1527,"a":1528},"Is there a Claude skills marketplace?","Anthropic publishes a set of official skills on GitHub, and Claude Code supports plugin marketplaces that bundle skills, commands and MCP servers. Several community directories exist. Treat skills like code: review what they run before installing.",{"q":1530,"a":1531},"Can I use Claude skills for free?","Yes. According to Anthropic, skills are available on the Free, Pro, Max, Team and Enterprise plans in Claude.ai, and they require code execution to be enabled. In Claude Code, skills are plain folders on disk and cost nothing beyond the tokens the agent uses when it reads them.",{"q":1533,"a":1534},"Is a Claude skill just a prompt?","Partly. The SKILL.md file is instructions, much like a prompt. The difference is that it loads only when relevant, it can carry scripts and reference files the agent runs or opens on demand, and it lives in a folder you can version, review and share. That makes it closer to a small package of know-how than to a pasted prompt.",{"q":1536,"a":1537},"Why is my Claude skill not being used?","Usually the description does not match how the request is phrased. In Claude Code, run \u002Fskills to check that the skill is loaded, invoke it directly with \u002Fskill-name to test it, and make the description name the task and the words people actually use. Malformed YAML frontmatter and disable-model-invocation: true are the other common causes.",{},0,"\u002Fguides\u002Fskills","14 min read",{"title":474,"description":1512},"guides\u002Fskills\u002Findex","claude-skills-guide","tgW8HrDNR6aWTqCL5PIqdPO_3U8zVzxj9cmFICkwnqU",{"id":1547,"title":1548,"author":6,"body":1549,"date":1511,"description":1841,"extension":441,"faq":1842,"meta":1855,"navigation":462,"order":60,"path":1856,"readTime":1857,"seo":1858,"stem":1859,"topic":468,"translationId":1860,"updated":1511,"__hash__":1861},"guides\u002Fguides\u002Fskills\u002Fclaude-skills-vs-mcp.md","Claude skills vs MCP: what each one does and when to use which",{"type":8,"value":1550,"toc":1832},[1551,1554,1565,1569,1581,1595,1599,1701,1705,1711,1714,1717,1720,1724,1727,1734,1762,1765,1769,1782,1785,1796,1799,1803,1806,1810,1813,1820],[11,1552,1553],{},"Two features arrived a year apart and get confused constantly. MCP, the Model Context Protocol, launched in November 2024. Skills launched in October 2025. Both extend what an agent can do, and both live in folders and config files, so it is natural to ask which one to use.",[11,1555,1556,1557,1560,1561,1564],{},"The short answer: MCP is about ",[127,1558,1559],{},"access",", skills are about ",[127,1562,1563],{},"know-how",". The rest of this page makes that precise.",[39,1566,1568],{"id":1567},"one-sentence-each","One sentence each",[11,1570,1571,1572,1575,1576,1580],{},"An ",[127,1573,1574],{},"MCP server"," gives the agent tools it did not have: search GitHub, query a database, post to Slack, drive a browser. It is a program running somewhere, exposing functions through a standard protocol. Read ",[18,1577,1579],{"href":1578},"\u002Fen\u002Fguides\u002Fmcp","our MCP guide"," for the full picture.",[11,1582,1583,1584,1587,1588,1590,1591,1594],{},"A ",[127,1585,1586],{},"skill"," gives the agent instructions it did not have: how to write release notes the way your team does, how to produce a valid Excel file, how to run your incident checklist. It is a folder with a ",[54,1589,482],{}," and optional scripts. Read ",[18,1592,1593],{"href":20},"our skills guide"," for the details, including where skills live in Claude Code, Claude.ai, the API and Managed Agents.",[39,1596,1598],{"id":1597},"side-by-side","Side by side",[179,1600,1601,1611],{},[182,1602,1603],{},[185,1604,1605,1607,1609],{},[188,1606],{},[188,1608,1574],{},[188,1610,190],{},[195,1612,1613,1624,1635,1646,1657,1668,1679,1690],{},[185,1614,1615,1618,1621],{},[200,1616,1617],{},"What it adds",[200,1619,1620],{},"Tools, resources, prompts from an external system",[200,1622,1623],{},"Instructions, examples, scripts for a task",[185,1625,1626,1629,1632],{},[200,1627,1628],{},"Where it runs",[200,1630,1631],{},"A separate process, local or remote",[200,1633,1634],{},"Inside the agent's context; scripts run in the sandbox or shell",[185,1636,1637,1640,1643],{},[200,1638,1639],{},"Needs credentials",[200,1641,1642],{},"Usually (OAuth, API keys)",[200,1644,1645],{},"Rarely; uses whatever tools are already available",[185,1647,1648,1651,1654],{},[200,1649,1650],{},"Loaded when",[200,1652,1653],{},"Every session (tool descriptions)",[200,1655,1656],{},"Only when a task matches its description",[185,1658,1659,1662,1665],{},[200,1660,1661],{},"Token cost",[200,1663,1664],{},"Per session, proportional to number of tools",[200,1666,1667],{},"Zero until used, then the file length",[185,1669,1670,1673,1676],{},[200,1671,1672],{},"Portable to other agents",[200,1674,1675],{},"Yes, any MCP client",[200,1677,1678],{},"Yes, via the open Agent Skills spec",[185,1680,1681,1684,1687],{},[200,1682,1683],{},"Main risk",[200,1685,1686],{},"Over-permission, prompt injection via results",[200,1688,1689],{},"Untrusted instructions and scripts",[185,1691,1692,1695,1698],{},[200,1693,1694],{},"Owned by",[200,1696,1697],{},"Usually the vendor of the system",[200,1699,1700],{},"Usually your team",[39,1702,1704],{"id":1703},"the-question-that-decides-it","The question that decides it",[11,1706,1707,1708],{},"Ask: ",[127,1709,1710],{},"does the agent lack access, or lack knowledge?",[11,1712,1713],{},"If the agent cannot do the task because it cannot reach something (the ticket system, the data warehouse, the deployment pipeline), that is an access problem. You need an MCP server, or a CLI the agent can call.",[11,1715,1716],{},"If the agent can reach everything but does the task badly (wrong format, wrong order, forgets a step, ignores a convention), that is a knowledge problem. You need a skill.",[11,1718,1719],{},"Most real tasks are both. \"Prepare the release\" needs access to GitHub and the deployment system (MCP) and the knowledge of what your release procedure is (skill).",[39,1721,1723],{"id":1722},"how-they-work-together","How they work together",[11,1725,1726],{},"Here is a typical workflow with both.",[11,1728,1729,1730,1733],{},"A developer types \"cut release 2.5\". The agent matches the request to a ",[54,1731,1732],{},"release"," skill and loads it. The skill says:",[121,1735,1736,1739,1745,1752,1755],{},[124,1737,1738],{},"Confirm the main branch is green using the CI tool.",[124,1740,1741,1742,27],{},"List merged PRs since the last tag using the GitHub tool, excluding ",[54,1743,1744],{},"chore",[124,1746,1747,1748,1751],{},"Draft release notes in the house format (example in ",[54,1749,1750],{},"examples\u002F",").",[124,1753,1754],{},"Create the tag and the GitHub release using the GitHub tool.",[124,1756,1757,1758,1761],{},"Post the notes to ",[54,1759,1760],{},"#releases"," using the Slack tool, then stop. Do not deploy.",[11,1763,1764],{},"Steps 1, 2, 4 and 5 call MCP tools. Steps 3 and the overall order come from the skill. Neither alone gets you a release done your way.",[39,1766,1768],{"id":1767},"when-a-skill-is-enough","When a skill is enough",[11,1770,1771,1772,1239,1775,148,1778,1781],{},"Skills can carry scripts, and scripts can call CLIs. If ",[54,1773,1774],{},"gh",[54,1776,1777],{},"aws",[54,1779,1780],{},"kubectl"," are already installed and authenticated on the machine, a skill can use them through the shell without any MCP server. For personal use that is often simpler.",[11,1783,1784],{},"It stops being enough when:",[342,1786,1787,1790,1793],{},[124,1788,1789],{},"the agent runs somewhere without those CLIs (Claude.ai, the API, a CI runner);",[124,1791,1792],{},"you need per-user authorization rather than whatever the laptop's credentials allow;",[124,1794,1795],{},"you want the access to be discoverable and governable across a team.",[11,1797,1798],{},"At that point the access belongs in an MCP server.",[39,1800,1802],{"id":1801},"when-mcp-is-enough","When MCP is enough",[11,1804,1805],{},"If the task is genuinely \"do the obvious thing with this tool\" (\"find open bugs assigned to me\", \"what changed in this file\"), tool descriptions carry enough guidance and no skill is needed. Skills earn their place when there is a procedure, a format or a convention the model would not guess.",[39,1807,1809],{"id":1808},"governance-they-meet-at-the-gateway","Governance: they meet at the gateway",[11,1811,1812],{},"For a team, both need an approved list. Skills are code and instructions from somewhere; MCP servers are code with credentials. A reviewed, versioned catalogue of each is the minimum.",[11,1814,1815,1816,1819],{},"But there is an asymmetry worth understanding. A skill describes a procedure; it cannot enforce who may run it. An MCP tool actually performs the action; it can be gated per user. So the enforcement point for \"only the release manager may tag a release\" is not the release skill but the policy on the GitHub server's ",[54,1817,1818],{},"create_release"," tool.",[11,1821,1822,1823,1827,1828,1831],{},"That is why, in a company deployment, both end up behind the same ",[18,1824,1826],{"href":1825},"\u002Fen\u002Fguides\u002Fmcp\u002Fmcp-gateway","gateway",": skills and servers are distributed through one client, tool policy is enforced per user, and the log shows which skill drove which tool call. Walma AI Hub does exactly that, inside the customer's own EU region. ",[18,1829,1830],{"href":424},"Book a walkthrough"," if you are working out how to standardise both across a team.",{"title":52,"searchDepth":95,"depth":431,"links":1833},[1834,1835,1836,1837,1838,1839,1840],{"id":1567,"depth":95,"text":1568},{"id":1597,"depth":95,"text":1598},{"id":1703,"depth":95,"text":1704},{"id":1722,"depth":95,"text":1723},{"id":1767,"depth":95,"text":1768},{"id":1801,"depth":95,"text":1802},{"id":1808,"depth":95,"text":1809},"Skills teach an agent how to do a task. MCP gives it access to tools and data. They are not competitors. Here is a clear comparison, the questions that decide which you need, and how they work together in a real workflow.",[1843,1846,1849,1852],{"q":1844,"a":1845},"Should I use skills or MCP?","Usually both. Use MCP when the agent needs to reach a system it cannot otherwise access, such as GitHub, a database or a SaaS tool. Use a skill when the agent needs to know how to do a task well, such as your team's review process or a document format. A skill often uses MCP tools to do its work.",{"q":1847,"a":1848},"Can a skill replace an MCP server?","Only if the task needs no external access, or the access is available through the shell or a CLI already on the machine. A skill can wrap a CLI in a script. For anything that needs authentication to a remote system, an MCP server is the cleaner and safer option.",{"q":1850,"a":1851},"Which is more secure, skills or MCP?","They have different risks. MCP servers hold credentials and can act on systems, so their risk is over-permission and injection through results. Skills run scripts and shape the model's behaviour, so their risk is supply chain: an untrusted skill can instruct the agent to do harmful things. Both need review and an allowlist in a team setting.",{"q":1853,"a":1854},"Do skills use more tokens than MCP?","A skill costs nothing until it is loaded, then costs the length of its SKILL.md. An MCP server costs its tool descriptions on every session. A server with 40 tools can be more expensive per session than several skills.",{},"\u002Fguides\u002Fskills\u002Fclaude-skills-vs-mcp","7 min read",{"title":1548,"description":1841},"guides\u002Fskills\u002Fclaude-skills-vs-mcp","claude-skills-vs-mcp","ddouen3aLaWj4u-zqv3xJGCkw974UrfCaoBq3ItsEXk",{"id":1863,"title":1864,"author":6,"body":1865,"date":2507,"description":2508,"extension":441,"faq":2509,"meta":2525,"navigation":462,"order":95,"path":2526,"readTime":1857,"seo":2527,"stem":2528,"topic":468,"translationId":469,"updated":2507,"__hash__":2529},"guides\u002Fguides\u002Fskills\u002Fanthropic-skills.md","Anthropic skills: the official GitHub repo, the document skills and which ones to install",{"type":8,"value":1866,"toc":2497},[1867,1870,1880,1884,1893,1924,1927,1931,1937,2012,2018,2022,2032,2053,2071,2085,2106,2109,2113,2124,2158,2175,2188,2192,2198,2208,2384,2390,2394,2397,2450,2454,2457,2464,2470,2474,2477,2494],[11,1868,1869],{},"When people search for \"Anthropic skills\" they usually mean one of two things: the public GitHub repository where Anthropic publishes its own skills, or the four document skills that let Claude produce real Word, Excel, PowerPoint and PDF files. They are the same thing seen from two sides. This guide covers what is actually in the official collection, how to install it in each product, which skills earn a place in a team setup, and what the licences let you do.",[11,1871,1872,1873,1876,1877,1879],{},"If you are new to skills as a concept, start with the ",[18,1874,1875],{"href":20},"Claude skills pillar guide",", which covers how a skill works and how to write one. If you are deciding between a skill and an MCP connection, read ",[18,1878,490],{"href":489},". This page sticks to Anthropic's own collection.",[39,1881,1883],{"id":1882},"the-anthropicsskills-repository","The anthropics\u002Fskills repository",[11,1885,1886,1887,1892],{},"The official repo is ",[18,1888,1891],{"href":1889,"rel":1890},"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fskills",[35],"github.com\u002Fanthropics\u002Fskills",", described as the \"Public repository for Agent Skills\" and actively maintained. The top level has four things worth knowing about:",[342,1894,1895,1903,1909,1915],{},[124,1896,1897,1900,1901,27],{},[54,1898,1899],{},"skills\u002F"," holds the skills themselves, one folder each with a ",[54,1902,482],{},[124,1904,1905,1908],{},[54,1906,1907],{},"spec\u002F"," holds the Agent Skills specification. The open standard itself lives at agentskills.io.",[124,1910,1911,1914],{},[54,1912,1913],{},"template\u002F"," is a minimal starter skill.",[124,1916,1917,1920,1921,27],{},[54,1918,1919],{},".claude-plugin\u002Fmarketplace.json"," turns the repo into a Claude Code plugin marketplace called ",[54,1922,1923],{},"anthropic-agent-skills",[11,1925,1926],{},"The README is explicit that these skills are \"provided for demonstration and educational purposes only\" and that what Claude does in its own products may differ from the code shown. Take that seriously: treat the repo as a set of well-built reference implementations, not as a supported product with an SLA.",[39,1928,1930],{"id":1929},"what-is-in-the-collection","What is in the collection",[11,1932,1933,1934,1936],{},"As of this writing the ",[54,1935,1899],{}," folder contains 19 skills. The marketplace groups them into five plugins:",[179,1938,1939,1951],{},[182,1940,1941],{},[185,1942,1943,1945,1948],{},[188,1944,841],{},[188,1946,1947],{},"Skills",[188,1949,1950],{},"Licence",[195,1952,1953,1966,1979,1990,2001],{},[185,1954,1955,1960,1963],{},[200,1956,1957],{},[54,1958,1959],{},"document-skills",[200,1961,1962],{},"docx, xlsx, pptx, pdf",[200,1964,1965],{},"Proprietary, source-available",[185,1967,1968,1973,1976],{},[200,1969,1970],{},[54,1971,1972],{},"example-skills",[200,1974,1975],{},"algorithmic-art, brand-guidelines, canvas-design, doc-coauthoring, frontend-design, internal-comms, mcp-builder, skill-creator, slack-gif-creator, theme-factory, web-artifacts-builder, webapp-testing",[200,1977,1978],{},"Apache 2.0",[185,1980,1981,1986,1988],{},[200,1982,1983],{},[54,1984,1985],{},"claude-api",[200,1987,1985],{},[200,1989,1978],{},[185,1991,1992,1997,1999],{},[200,1993,1994],{},[54,1995,1996],{},"academy-guide",[200,1998,1996],{},[200,2000,1978],{},[185,2002,2003,2008,2010],{},[200,2004,2005],{},[54,2006,2007],{},"discernment-nudge",[200,2009,2007],{},[200,2011,1978],{},[11,2013,2014,2015,2017],{},"The split matters. ",[54,2016,1972],{}," is a grab bag: some entries are genuinely useful at work, others (algorithmic art, Slack GIFs) exist to show what the format can do. The document skills are a different category entirely. They are the same skills that power Claude's file creation in claude.ai, published so developers can see how a complex production skill is structured.",[39,2019,2021],{"id":2020},"the-document-skills-docx-xlsx-pptx-pdf","The document skills: docx, xlsx, pptx, pdf",[11,2023,2024,2025,2027,2028,2031],{},"Each document skill is a ",[54,2026,482],{}," plus a ",[54,2029,2030],{},"scripts\u002F"," folder of helpers.",[11,2033,2034,2037,2038,2040,2041,2044,2045,2048,2049,2052],{},[127,2035,2036],{},"docx"," creates new Word documents with the ",[54,2039,2036],{}," npm library, and edits existing ones by unzipping the file and changing ",[54,2042,2043],{},"word\u002Fdocument.xml"," directly, since that library cannot open existing files. It reads documents with ",[54,2046,2047],{},"pandoc",", handles tracked changes and comments, and ships an ",[54,2050,2051],{},"accept_changes.py"," script for producing a clean copy.",[11,2054,2055,2058,2059,2062,2063,2066,2067,2070],{},[127,2056,2057],{},"xlsx"," uses ",[54,2060,2061],{},"openpyxl"," for formulas and formatting and ",[54,2064,2065],{},"pandas"," for bulk data. Its hard rules are the useful part: use formulas rather than hardcoded results, and never ship a file with formula errors. Because openpyxl writes formulas without cached values, the skill runs a ",[54,2068,2069],{},"recalc.py"," script that recalculates every formula through LibreOffice and reports errors as JSON.",[11,2072,2073,2076,2077,2080,2081,2084],{},[127,2074,2075],{},"pptx"," builds decks with ",[54,2078,2079],{},"pptxgenjs",", reads them with ",[54,2082,2083],{},"markitdown",", and renders slides to images through LibreOffice and Poppler so Claude can check the visual result instead of trusting the XML.",[11,2086,2087,2090,2091,915,2094,2097,2098,2101,2102,2105],{},[127,2088,2089],{},"pdf"," covers extraction with ",[54,2092,2093],{},"pdfplumber",[54,2095,2096],{},"pypdf",", creation with ReportLab, merging and splitting with ",[54,2099,2100],{},"qpdf",", OCR with ",[54,2103,2104],{},"pytesseract",", and a set of scripts for filling both fillable and non-fillable forms.",[11,2107,2108],{},"The practical consequence for Claude Code users: these skills expect tools on the machine. In claude.ai and the API they run in Anthropic's sandbox, where the dependencies are preinstalled. On your laptop, the docx, pptx and xlsx workflows lean on LibreOffice, Poppler and pandoc, and the PDF skill on qpdf and Tesseract for OCR. Install what is missing before you judge the skill.",[39,2110,2112],{"id":2111},"installing-in-claude-code","Installing in Claude Code",[11,2114,2115,2116,2119,2120,2123],{},"Claude Code reads skills from ",[54,2117,2118],{},"~\u002F.claude\u002Fskills\u002F"," (personal) and ",[54,2121,2122],{},".claude\u002Fskills\u002F"," (per project), and from installed plugins. Anthropic's own docs note that the pre-built document skills are not part of Claude Code out of the box; only the Claude API skill comes bundled. The repo's marketplace fills that gap. Inside a Claude Code session:",[47,2125,2127],{"className":49,"code":2126,"language":51,"meta":52,"style":52},"\u002Fplugin marketplace add anthropics\u002Fskills\n\u002Fplugin install document-skills@anthropic-agent-skills\n\u002Fplugin install example-skills@anthropic-agent-skills\n",[54,2128,2129,2140,2149],{"__ignoreMap":52},[57,2130,2131,2133,2135,2137],{"class":59,"line":60},[57,2132,64],{"class":63},[57,2134,86],{"class":67},[57,2136,89],{"class":67},[57,2138,2139],{"class":67}," anthropics\u002Fskills\n",[57,2141,2142,2144,2146],{"class":59,"line":95},[57,2143,64],{"class":63},[57,2145,68],{"class":67},[57,2147,2148],{"class":67}," document-skills@anthropic-agent-skills\n",[57,2150,2151,2153,2155],{"class":59,"line":431},[57,2152,64],{"class":63},[57,2154,68],{"class":67},[57,2156,2157],{"class":67}," example-skills@anthropic-agent-skills\n",[11,2159,2160,2161,2163,2164,2166,2167,2170,2171,2174],{},"You can also run ",[54,2162,64],{},", choose \"Browse and install plugins\", pick ",[54,2165,1923],{}," and then the plugin you want. After that, Claude picks a skill up automatically when your request matches its description (\"turn this CSV into a formatted Excel workbook\"), or you can invoke it by name. Plugin skills are namespaced, so the PDF skill is reachable as ",[54,2168,2169],{},"\u002Fdocument-skills:pdf",", and usually by the bare ",[54,2172,2173],{},"\u002Fpdf"," if nothing else claims that name.",[11,2176,2177,2178,2180,2181,2183,2184,27],{},"If you only want one skill from ",[54,2179,1972],{},", the plugin is all or nothing. The Apache-licensed skills can instead be copied folder by folder into ",[54,2182,2122],{}," and committed with the project. Do not do that with the document skills; see the licence section below. For more on setting up Claude Code itself, see ",[18,2185,2187],{"href":2186},"\u002Fen\u002Fguides\u002Fclaude-code\u002Fhow-to-install-claude-code","how to install Claude Code",[39,2189,2191],{"id":2190},"using-them-in-claudeai-and-the-api","Using them in Claude.ai and the API",[11,2193,2194,2197],{},[127,2195,2196],{},"Claude.ai."," You do not install the document skills here. They are used automatically when Claude creates files, provided code execution is on. On individual plans that is Settings, Capabilities; on Team and Enterprise an owner first enables code execution and Skills in the organisation settings. Skills from the repo beyond the document set are, according to the README, already available on paid plans. To add your own, zip the skill folder and upload it under Customize, Skills.",[11,2199,2200,2203,2204,2207],{},[127,2201,2202],{},"Claude API."," Attach pre-built skills by ID in the ",[54,2205,2206],{},"container"," parameter and include the code execution tool, whose sandbox the skills run in:",[47,2209,2213],{"className":2210,"code":2211,"language":2212,"meta":52,"style":52},"language-json shiki shiki-themes github-dark","{\n  \"model\": \"claude-opus-5-5\",\n  \"max_tokens\": 4096,\n  \"container\": {\n    \"skills\": [\n      { \"type\": \"anthropic\", \"skill_id\": \"xlsx\", \"version\": \"latest\" }\n    ]\n  },\n  \"tools\": [\n    { \"type\": \"code_execution_20250825\", \"name\": \"code_execution\" }\n  ],\n  \"messages\": [\n    { \"role\": \"user\", \"content\": \"Build a 12-month budget workbook with formulas.\" }\n  ]\n}\n","json",[54,2214,2215,2220,2233,2245,2253,2261,2297,2302,2307,2314,2338,2343,2350,2374,2379],{"__ignoreMap":52},[57,2216,2217],{"class":59,"line":60},[57,2218,2219],{"class":572},"{\n",[57,2221,2222,2225,2227,2230],{"class":59,"line":95},[57,2223,2224],{"class":635},"  \"model\"",[57,2226,1031],{"class":572},[57,2228,2229],{"class":67},"\"claude-opus-5-5\"",[57,2231,2232],{"class":572},",\n",[57,2234,2235,2238,2240,2243],{"class":59,"line":431},[57,2236,2237],{"class":635},"  \"max_tokens\"",[57,2239,1031],{"class":572},[57,2241,2242],{"class":635},"4096",[57,2244,2232],{"class":572},[57,2246,2247,2250],{"class":59,"line":581},[57,2248,2249],{"class":635},"  \"container\"",[57,2251,2252],{"class":572},": {\n",[57,2254,2255,2258],{"class":59,"line":586},[57,2256,2257],{"class":635},"    \"skills\"",[57,2259,2260],{"class":572},": [\n",[57,2262,2263,2266,2269,2271,2274,2276,2279,2281,2284,2286,2289,2291,2294],{"class":59,"line":463},[57,2264,2265],{"class":572},"      { ",[57,2267,2268],{"class":635},"\"type\"",[57,2270,1031],{"class":572},[57,2272,2273],{"class":67},"\"anthropic\"",[57,2275,1239],{"class":572},[57,2277,2278],{"class":635},"\"skill_id\"",[57,2280,1031],{"class":572},[57,2282,2283],{"class":67},"\"xlsx\"",[57,2285,1239],{"class":572},[57,2287,2288],{"class":635},"\"version\"",[57,2290,1031],{"class":572},[57,2292,2293],{"class":67},"\"latest\"",[57,2295,2296],{"class":572}," }\n",[57,2298,2299],{"class":59,"line":597},[57,2300,2301],{"class":572},"    ]\n",[57,2303,2304],{"class":59,"line":602},[57,2305,2306],{"class":572},"  },\n",[57,2308,2309,2312],{"class":59,"line":608},[57,2310,2311],{"class":635},"  \"tools\"",[57,2313,2260],{"class":572},[57,2315,2316,2319,2321,2323,2326,2328,2331,2333,2336],{"class":59,"line":614},[57,2317,2318],{"class":572},"    { ",[57,2320,2268],{"class":635},[57,2322,1031],{"class":572},[57,2324,2325],{"class":67},"\"code_execution_20250825\"",[57,2327,1239],{"class":572},[57,2329,2330],{"class":635},"\"name\"",[57,2332,1031],{"class":572},[57,2334,2335],{"class":67},"\"code_execution\"",[57,2337,2296],{"class":572},[57,2339,2340],{"class":59,"line":619},[57,2341,2342],{"class":572},"  ],\n",[57,2344,2345,2348],{"class":59,"line":625},[57,2346,2347],{"class":635},"  \"messages\"",[57,2349,2260],{"class":572},[57,2351,2352,2354,2357,2359,2362,2364,2367,2369,2372],{"class":59,"line":647},[57,2353,2318],{"class":572},[57,2355,2356],{"class":635},"\"role\"",[57,2358,1031],{"class":572},[57,2360,2361],{"class":67},"\"user\"",[57,2363,1239],{"class":572},[57,2365,2366],{"class":635},"\"content\"",[57,2368,1031],{"class":572},[57,2370,2371],{"class":67},"\"Build a 12-month budget workbook with formulas.\"",[57,2373,2296],{"class":572},[57,2375,2376],{"class":59,"line":656},[57,2377,2378],{"class":572},"  ]\n",[57,2380,2381],{"class":59,"line":665},[57,2382,2383],{"class":572},"}\n",[11,2385,2386,2387,2389],{},"Up to 20 skills can be attached per request. API skills run with no network access and no runtime package installation, which is the right default for anything touching business data. Custom skills uploaded through ",[54,2388,1127],{}," are shared across the workspace. Note that skills do not sync between surfaces: a skill uploaded to claude.ai is not available in the API or in Claude Code, and the other way round.",[39,2391,2393],{"id":2392},"which-ones-are-worth-using","Which ones are worth using",[11,2395,2396],{},"Our opinion after using the collection on real work:",[342,2398,2399,2409,2415,2426,2432,2438,2444],{},[124,2400,2401,2404,2405,27],{},[127,2402,2403],{},"document-skills: yes."," The single most useful thing in the repo. If your team generates reports, board decks or finance workbooks from Claude Code, this is what stops broken formulas and slides that look fine in XML and terrible on screen. The xlsx skill alone justifies it for anyone in ",[18,2406,2408],{"href":2407},"\u002Fen\u002Fai-for-finance","finance and reporting",[124,2410,2411,2414],{},[127,2412,2413],{},"skill-creator: yes."," It walks through writing a skill, runs evals against it and helps tune the description so the skill triggers when it should. Use it before you write your team's first internal skill.",[124,2416,2417,2420,2421,2425],{},[127,2418,2419],{},"mcp-builder: yes, for engineers."," A solid guide for building MCP servers in Python (FastMCP) or TypeScript. Pair it with our ",[18,2422,2424],{"href":2423},"\u002Fen\u002Fguides\u002Fmcp\u002Fmcp-security-best-practices","MCP security best practices"," before anything goes near production data.",[124,2427,2428,2431],{},[127,2429,2430],{},"webapp-testing and frontend-design: worth a try"," in front-end projects. The first drives Playwright to check local web apps; the second pushes Claude away from generic-looking UI.",[124,2433,2434,2437],{},[127,2435,2436],{},"claude-api: already there"," if you use Claude Code, since it ships bundled.",[124,2439,2440,2443],{},[127,2441,2442],{},"doc-coauthoring and internal-comms: situational."," Useful structure for specs and status updates, but most teams will want their own versions with house style.",[124,2445,2446,2449],{},[127,2447,2448],{},"The rest:"," fun, instructive, rarely needed at work.",[39,2451,2453],{"id":2452},"licence-notes","Licence notes",[11,2455,2456],{},"Read the licences before you fork or redistribute anything.",[11,2458,2459,2460,2463],{},"Most skills carry an Apache 2.0 ",[54,2461,2462],{},"LICENSE.txt",". You can copy, modify and ship them, including inside a company plugin, as long as you keep the licence and notices.",[11,2465,2466,2467,2469],{},"The four document skills are different. Their ",[54,2468,2462],{}," says use is governed by your agreement with Anthropic and, among other restrictions, forbids reproducing the materials, creating derivative works, and distributing them to third parties. In practice: installing them through the official marketplace and using them is what Anthropic intends; vendoring a modified copy of the xlsx skill into your internal skill pack is not. If you need a house-style variant, write your own skill that sits alongside the official one rather than editing it.",[39,2471,2473],{"id":2472},"governance-skills-are-code-treat-them-like-code","Governance: skills are code, treat them like code",[11,2475,2476],{},"Anthropic's documentation is blunt about this: use skills only from trusted sources, because a skill can direct Claude to run code and call tools in ways that do not match its stated purpose. The official repo is about as trusted as it gets, but the same discipline should apply to every community skill someone on your team finds: review the scripts, pin a version, and know what data the agent can reach while it runs.",[11,2478,2479,2480,2483,2484,2488,2489,2493],{},"Two details matter for EU teams. First, Anthropic states that Agent Skills is not covered by zero data retention arrangements; skill definitions and execution data follow its standard retention policy. Second, skills decide how Claude works, while MCP connections decide what it can reach. That second part is where most of the risk sits. At Walma we run that layer in the customer's own Azure tenant in an EU region: every model, the MCP connections, policy, cost control and logging in one place. See the ",[18,2481,2482],{"href":424},"AI Hub"," for how it fits together, or the ",[18,2485,2487],{"href":2486},"\u002Fen\u002Fguides\u002Fclaude-code","Claude Code guide"," for running the agent itself under policy. If an installed skill never seems to fire, the ",[18,2490,2492],{"href":2491},"\u002Fen\u002Fguides\u002Fskills#when-a-skill-does-not-trigger","troubleshooting checklist in our skills guide"," covers the usual causes.",[427,2495,2496],{},"html pre.shiki code .svObZ, html code.shiki .svObZ{--shiki-default:#B392F0}html pre.shiki code .sU2Wk, html code.shiki .sU2Wk{--shiki-default:#9ECBFF}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 pre.shiki code .s95oV, html code.shiki .s95oV{--shiki-default:#E1E4E8}html pre.shiki code .sDLfK, html code.shiki .sDLfK{--shiki-default:#79B8FF}",{"title":52,"searchDepth":95,"depth":431,"links":2498},[2499,2500,2501,2502,2503,2504,2505,2506],{"id":1882,"depth":95,"text":1883},{"id":1929,"depth":95,"text":1930},{"id":2020,"depth":95,"text":2021},{"id":2111,"depth":95,"text":2112},{"id":2190,"depth":95,"text":2191},{"id":2392,"depth":95,"text":2393},{"id":2452,"depth":95,"text":2453},{"id":2472,"depth":95,"text":2473},"2026-10-05","A practical guide to Anthropic's official skills: what is in the anthropics\u002Fskills GitHub repository, how the docx, xlsx, pptx and pdf document skills work, how to install them in Claude Code through the plugin marketplace and use them in Claude.ai and the API, which ones are worth your time, and what the licences allow.",[2510,2513,2516,2519,2522],{"q":2511,"a":2512},"Where are Anthropic's official skills on GitHub?","In the public repository github.com\u002Fanthropics\u002Fskills. It holds the skills under skills\u002F, the Agent Skills specification under spec\u002F, a starter template under template\u002F, and a .claude-plugin\u002Fmarketplace.json that lets Claude Code install the skills as plugins.",{"q":2514,"a":2515},"How do I install Anthropic skills in Claude Code?","Inside Claude Code, run \u002Fplugin marketplace add anthropics\u002Fskills, then \u002Fplugin install document-skills@anthropic-agent-skills for the Word, Excel, PowerPoint and PDF skills, or \u002Fplugin install example-skills@anthropic-agent-skills for the rest. You can also browse and install from the \u002Fplugin menu.",{"q":2517,"a":2518},"Are Anthropic's skills open source?","Partly. Most skills in the repository are Apache 2.0. The four document skills (docx, xlsx, pptx, pdf) are source-available under a proprietary licence that ties use to your agreement with Anthropic and forbids copying, derivative works and redistribution.",{"q":2520,"a":2521},"Do I need to install the document skills in Claude.ai?","No. On claude.ai the pre-built document skills are used automatically when Claude creates files, as long as code execution and file creation is turned on. Installing them yourself is mainly relevant in Claude Code.",{"q":2523,"a":2524},"Can I use the document skills through the Claude API?","Yes. Reference them by skill_id (pptx, xlsx, docx or pdf) with type anthropic in the container parameter, together with the code execution tool. Up to 20 skills can be attached to one request.",{},"\u002Fguides\u002Fskills\u002Fanthropic-skills",{"title":1864,"description":2508},"guides\u002Fskills\u002Fanthropic-skills","EF43K2MC0d16vSwuj6xI5dQXWiu_1Gne7V_Jg2C8qLw",{"id":2531,"title":2532,"author":6,"body":2533,"date":3712,"description":3713,"extension":441,"faq":3714,"meta":3748,"navigation":462,"order":431,"path":3749,"readTime":1541,"seo":3750,"stem":3751,"topic":468,"translationId":469,"updated":439,"__hash__":3752},"guides\u002Fguides\u002Fskills\u002Fbest-claude-skills.md","The best Claude skills in 2026: top picks by use case",{"type":8,"value":2534,"toc":3696},[2535,2538,2555,2558,2562,2565,2794,2797,2801,2804,2824,2844,2848,2855,2885,2888,3029,3040,3044,3053,3065,3073,3092,3117,3129,3137,3141,3152,3177,3189,3193,3196,3205,3230,3239,3255,3269,3272,3283,3287,3290,3299,3324,3338,3341,3345,3354,3379,3388,3413,3434,3438,3448,3464,3475,3479,3482,3519,3531,3535,3538,3541,3568,3571,3575,3628,3632,3661,3663,3666,3686,3693],[11,2536,2537],{},"There are now thousands of Claude skills on GitHub. Most are prompts in a folder, some are excellent, and a few are worth installing on day one. This list is the short version: the skills that are widely used, actively maintained and do something a plain prompt cannot, grouped by the job they do.",[11,2539,2540,2541,2544,2545,2549,2550,2554],{},"If skills are new to you, start with ",[18,2542,2543],{"href":20},"what Claude skills are and how they work",". Every recommendation here comes with the same caveat: a skill runs with your permissions, so read it before you install it. The ",[18,2546,2548],{"href":2547},"\u002Fen\u002Fguides\u002Fskills#before-you-install-someone-elses-skill","review checklist"," takes five minutes, and ",[18,2551,2553],{"href":2552},"#check-a-skill-before-you-enable-it","the section below"," covers scanning.",[11,2556,2557],{},"Repositories, star counts and install commands were checked on GitHub on 12 October 2026.",[39,2559,2561],{"id":2560},"the-top-10-claude-skills-at-a-glance","The top 10 Claude skills at a glance",[11,2563,2564],{},"Ranked by GitHub stars, which measure attention rather than quality. Every one of these is maintained and does something a prompt alone cannot.",[179,2566,2567,2584],{},[182,2568,2569],{},[185,2570,2571,2574,2576,2578,2581],{},[188,2572,2573],{},"#",[188,2575,190],{},[188,2577,193],{},[188,2579,2580],{},"Stars",[188,2582,2583],{},"Best for",[195,2585,2586,2606,2626,2647,2668,2689,2710,2731,2752,2773],{},[185,2587,2588,2591,2597,2600,2603],{},[200,2589,2590],{},"1",[200,2592,2593],{},[18,2594,2596],{"href":33,"rel":2595},[35],"superpowers",[200,2598,2599],{},"Plan, test-first, debug systematically",[200,2601,2602],{},"~298k",[200,2604,2605],{},"Software teams",[185,2607,2608,2611,2617,2620,2623],{},[200,2609,2610],{},"2",[200,2612,2613],{},[18,2614,2616],{"href":1889,"rel":2615},[35],"anthropics\u002Fskills",[200,2618,2619],{},"Official: Word, Excel, PowerPoint, PDF, skill-creator",[200,2621,2622],{},"~181k",[200,2624,2625],{},"Everyone",[185,2627,2628,2631,2638,2641,2644],{},[200,2629,2630],{},"3",[200,2632,2633],{},[18,2634,2637],{"href":2635,"rel":2636},"https:\u002F\u002Fgithub.com\u002Fnextlevelbuilder\u002Fui-ux-pro-max-skill",[35],"ui-ux-pro-max",[200,2639,2640],{},"Design rules, palettes, fonts and UX checks for UI work",[200,2642,2643],{},"~135k",[200,2645,2646],{},"Websites and apps",[185,2648,2649,2652,2659,2662,2665],{},[200,2650,2651],{},"4",[200,2653,2654],{},[18,2655,2658],{"href":2656,"rel":2657},"https:\u002F\u002Fgithub.com\u002FJuliusBrussee\u002Fcaveman",[35],"caveman",[200,2660,2661],{},"Short answers, fewer tokens, code untouched",[200,2663,2664],{},"~111k",[200,2666,2667],{},"Cutting token cost",[185,2669,2670,2673,2680,2683,2686],{},[200,2671,2672],{},"5",[200,2674,2675],{},[18,2676,2679],{"href":2677,"rel":2678},"https:\u002F\u002Fgithub.com\u002FLeonxlnx\u002Ftaste-skill",[35],"taste-skill",[200,2681,2682],{},"Pushes UI away from generic AI looks",[200,2684,2685],{},"~95k",[200,2687,2688],{},"Front end",[185,2690,2691,2694,2701,2704,2707],{},[200,2692,2693],{},"6",[200,2695,2696],{},[18,2697,2700],{"href":2698,"rel":2699},"https:\u002F\u002Fgithub.com\u002Ftt-a1i\u002Farchify",[35],"archify",[200,2702,2703],{},"Turns a plan or codebase into an interactive diagram",[200,2705,2706],{},"~82k",[200,2708,2709],{},"Architecture and docs",[185,2711,2712,2715,2722,2725,2728],{},[200,2713,2714],{},"7",[200,2716,2717],{},[18,2718,2721],{"href":2719,"rel":2720},"https:\u002F\u002Fgithub.com\u002Fmvanhorn\u002Flast30days-skill",[35],"last30days",[200,2723,2724],{},"Researches a topic across Reddit, X, YouTube, HN",[200,2726,2727],{},"~64k",[200,2729,2730],{},"Research and trend briefs",[185,2732,2733,2736,2743,2746,2749],{},[200,2734,2735],{},"8",[200,2737,2738],{},[18,2739,2742],{"href":2740,"rel":2741},"https:\u002F\u002Fgithub.com\u002Fblader\u002Fhumanizer",[35],"humanizer",[200,2744,2745],{},"Removes AI tells from text",[200,2747,2748],{},"~56k",[200,2750,2751],{},"Writing",[185,2753,2754,2757,2764,2767,2770],{},[200,2755,2756],{},"9",[200,2758,2759],{},[18,2760,2763],{"href":2761,"rel":2762},"https:\u002F\u002Fgithub.com\u002Fcoreyhaines31\u002Fmarketingskills",[35],"marketingskills",[200,2765,2766],{},"~50 skills: copy, CRO, SEO, ads, analytics",[200,2768,2769],{},"~54k",[200,2771,2772],{},"Marketing",[185,2774,2775,2778,2785,2788,2791],{},[200,2776,2777],{},"10",[200,2779,2780],{},[18,2781,2784],{"href":2782,"rel":2783},"https:\u002F\u002Fgithub.com\u002Fkepano\u002Fobsidian-skills",[35],"obsidian-skills",[200,2786,2787],{},"Teaches the agent Obsidian's CLI and formats",[200,2789,2790],{},"~49k",[200,2792,2793],{},"Notes and knowledge bases",[11,2795,2796],{},"The rest of the page goes through these by job, with install commands and the runners-up worth knowing.",[39,2798,2800],{"id":2799},"how-we-picked","How we picked",[11,2802,2803],{},"A skill made the list if it is:",[342,2805,2806,2812,2818],{},[124,2807,2808,2811],{},[127,2809,2810],{},"Widely used",": among the most-starred skills or collections on GitHub, or official from Anthropic.",[124,2813,2814,2817],{},[127,2815,2816],{},"More than a prompt",": it carries scripts, references or a procedure that changes results.",[124,2819,2820,2823],{},[127,2821,2822],{},"Maintained",": recent commits and a clear install path.",[11,2825,2826,2827,2832,2833,2835,2836,2838,2839,2843],{},"Not every popular \"skills\" repository is a skill. ",[18,2828,2831],{"href":2829,"rel":2830},"https:\u002F\u002Fgithub.com\u002Fmultica-ai\u002Fandrej-karpathy-skills",[35],"andrej-karpathy-skills",", one of the most-starred, is a single ",[54,2834,763],{}," file of coding guidelines. Useful, but it belongs in your project's ",[54,2837,763],{},", not in your skills folder. The ",[18,2840,2842],{"href":2841},"\u002Fen\u002Fguides\u002Fskills#skills-vs-the-alternatives","skills guide"," explains the difference.",[39,2845,2847],{"id":2846},"official-skills-from-anthropic","Official skills from Anthropic",[11,2849,2850,2851,2854],{},"Anthropic's ",[18,2852,2616],{"href":1889,"rel":2853},[35]," repository is the reference collection. In Claude Code:",[47,2856,2857],{"className":49,"code":2126,"language":51,"meta":52,"style":52},[54,2858,2859,2869,2877],{"__ignoreMap":52},[57,2860,2861,2863,2865,2867],{"class":59,"line":60},[57,2862,64],{"class":63},[57,2864,86],{"class":67},[57,2866,89],{"class":67},[57,2868,2139],{"class":67},[57,2870,2871,2873,2875],{"class":59,"line":95},[57,2872,64],{"class":63},[57,2874,68],{"class":67},[57,2876,2148],{"class":67},[57,2878,2879,2881,2883],{"class":59,"line":431},[57,2880,64],{"class":63},[57,2882,68],{"class":67},[57,2884,2157],{"class":67},[11,2886,2887],{},"In Claude.ai, these are already available; enable them under the Skills settings.",[179,2889,2890,2900],{},[182,2891,2892],{},[185,2893,2894,2896,2898],{},[188,2895,190],{},[188,2897,193],{},[188,2899,2583],{},[195,2901,2902,2914,2929,2942,2955,2968,2981,2994,3010],{},[185,2903,2904,2908,2911],{},[200,2905,2906],{},[127,2907,1962],{},[200,2909,2910],{},"Create, read and edit Word, Excel, PowerPoint and PDF files properly: real formulas, real slides, filled forms",[200,2912,2913],{},"Anyone producing documents. The single most useful set.",[185,2915,2916,2920,2923],{},[200,2917,2918],{},[127,2919,1419],{},[200,2921,2922],{},"Builds new skills and tests them against a baseline with evals",[200,2924,2925,2926,27],{},"Writing your own skills. See our ",[18,2927,2928],{"href":1418},"skill-creator guide",[185,2930,2931,2936,2939],{},[200,2932,2933],{},[127,2934,2935],{},"frontend-design",[200,2937,2938],{},"Pushes UI work away from generic AI defaults toward deliberate typography, colour and layout",[200,2940,2941],{},"Developers and designers building interfaces",[185,2943,2944,2949,2952],{},[200,2945,2946],{},[127,2947,2948],{},"mcp-builder",[200,2950,2951],{},"Guides building an MCP server in Python (FastMCP) or TypeScript",[200,2953,2954],{},"Teams connecting internal systems",[185,2956,2957,2962,2965],{},[200,2958,2959],{},[127,2960,2961],{},"webapp-testing",[200,2963,2964],{},"Tests local web apps with Playwright, with screenshots and logs",[200,2966,2967],{},"Front-end QA",[185,2969,2970,2975,2978],{},[200,2971,2972],{},[127,2973,2974],{},"internal-comms",[200,2976,2977],{},"Status reports, leadership updates, newsletters and FAQs in a company's formats",[200,2979,2980],{},"Managers and internal communications",[185,2982,2983,2988,2991],{},[200,2984,2985],{},[127,2986,2987],{},"doc-coauthoring",[200,2989,2990],{},"A structured workflow for co-writing specs, proposals and docs",[200,2992,2993],{},"Product and engineering docs",[185,2995,2996,3004,3007],{},[200,2997,2998,1239,3001],{},[127,2999,3000],{},"theme-factory",[127,3002,3003],{},"brand-guidelines",[200,3005,3006],{},"Apply a colour and font theme, or a brand, to any artifact",[200,3008,3009],{},"Consistent-looking output",[185,3011,3012,3023,3026],{},[200,3013,3014,1239,3017,1239,3020],{},[127,3015,3016],{},"canvas-design",[127,3018,3019],{},"algorithmic-art",[127,3021,3022],{},"slack-gif-creator",[200,3024,3025],{},"Posters, generative art, Slack GIFs",[200,3027,3028],{},"Fun, and good examples to learn from",[11,3030,3031,3032,3035,3036,27],{},"Note the licences: most are Apache 2.0, but the four document skills are ",[127,3033,3034],{},"source-available under a proprietary licence",". You can use them with Claude; you cannot fork them into your own product. More detail in our ",[18,3037,3039],{"href":3038},"\u002Fen\u002Fguides\u002Fskills\u002Fanthropic-skills","Anthropic skills guide",[39,3041,3043],{"id":3042},"best-claude-skills-for-coding","Best Claude skills for coding",[11,3045,3046,3048,3049,3052],{},[127,3047,2596],{}," (",[18,3050,36],{"href":33,"rel":3051},[35],") is the most widely used skill collection for developers: brainstorming, writing plans, executing plans, test-driven development, systematic debugging, sub-agent-driven development and working with git worktrees. It is more a methodology than a single skill.",[47,3054,3055],{"className":49,"code":50,"language":51,"meta":52,"style":52},[54,3056,3057],{"__ignoreMap":52},[57,3058,3059,3061,3063],{"class":59,"line":60},[57,3060,64],{"class":63},[57,3062,68],{"class":67},[57,3064,71],{"class":67},[11,3066,3067,3068,3072],{},"If you adopt it, adopt it as a team. Its value is that every developer's agent works the same way, plans before code, tests before claiming done. Our ",[18,3069,3071],{"href":3070},"\u002Fen\u002Fguides\u002Fskills\u002Fclaude-superpowers","Superpowers guide"," walks through the workflow, every skill in it and what it costs in tokens.",[11,3074,3075,3048,3077,3081,3082,1239,3085,1239,3088,3091],{},[127,3076,2658],{},[18,3078,3080],{"href":2656,"rel":3079},[35],"JuliusBrussee\u002Fcaveman",") makes the agent answer in short, structured sentences while leaving code, commands, paths and error messages exactly as they are. The project's own example cuts a 63-token answer to 20. It comes in levels (",[54,3083,3084],{},"\u002Fcaveman",[54,3086,3087],{},"\u002Fultracave",[54,3089,3090],{},"\u002Fmegacave","); the first is the one to try.",[47,3093,3095],{"className":49,"code":3094,"language":51,"meta":52,"style":52},"\u002Fplugin marketplace add JuliusBrussee\u002Fcaveman\n\u002Fplugin install caveman@caveman\n",[54,3096,3097,3108],{"__ignoreMap":52},[57,3098,3099,3101,3103,3105],{"class":59,"line":60},[57,3100,64],{"class":63},[57,3102,86],{"class":67},[57,3104,89],{"class":67},[57,3106,3107],{"class":67}," JuliusBrussee\u002Fcaveman\n",[57,3109,3110,3112,3114],{"class":59,"line":95},[57,3111,64],{"class":63},[57,3113,68],{"class":67},[57,3115,3116],{"class":67}," caveman@caveman\n",[11,3118,3119,3048,3121,3125,3126,27],{},[127,3120,1212],{},[18,3122,3124],{"href":1210,"rel":3123},[35],"Walma-Labs\u002Fux-lint","), our own, is a UX linter that measures a page in a real browser (tap targets, layout shift, LCP, click-to-paint) and reasons from the numbers rather than guessing from source code. We use it as the worked example in our ",[18,3127,2842],{"href":3128},"\u002Fen\u002Fguides\u002Fskills#a-worked-example-ux-lint",[11,3130,3131,3132,915,3134,3136],{},"From Anthropic's set, ",[127,3133,2961],{},[127,3135,2948],{}," belong here too. Pick one methodology collection, superpowers or something like it, not several: two sets of rules about when to plan and when to test give the model conflicting orders.",[39,3138,3140],{"id":3139},"best-claude-skills-for-planning","Best Claude skills for planning",[11,3142,3143,3048,3146,3151],{},[127,3144,3145],{},"planning-with-files",[18,3147,3150],{"href":3148,"rel":3149},"https:\u002F\u002Fgithub.com\u002FOthmanAdi\u002Fplanning-with-files",[35],"OthmanAdi\u002Fplanning-with-files",") solves the most common planning problem in long tasks: an agent forgets its plan when the context is cleared or compacted. It keeps the plan, findings and progress in markdown files in the project and re-injects them every turn, so a long task resumes at the right phase. The project reports re-orientation after a reset dropping from about 13 turns to 5.",[47,3153,3155],{"className":49,"code":3154,"language":51,"meta":52,"style":52},"npx skills add OthmanAdi\u002Fplanning-with-files --skill planning-with-files -g\n",[54,3156,3157],{"__ignoreMap":52},[57,3158,3159,3161,3163,3165,3168,3171,3174],{"class":59,"line":60},[57,3160,1223],{"class":63},[57,3162,1226],{"class":67},[57,3164,89],{"class":67},[57,3166,3167],{"class":67}," OthmanAdi\u002Fplanning-with-files",[57,3169,3170],{"class":635}," --skill",[57,3172,3173],{"class":67}," planning-with-files",[57,3175,3176],{"class":635}," -g\n",[11,3178,3179,3182,3183,3188],{},[127,3180,3181],{},"llm-council"," style skills run several sub-agents as \"advisors\" who argue, review each other and reach a verdict, inspired by Andrej Karpathy's LLM Council experiment. ",[18,3184,3187],{"href":3185,"rel":3186},"https:\u002F\u002Fgithub.com\u002Ftenfoldmarc\u002Fllm-council-skill",[35],"tenfoldmarc\u002Fllm-council-skill"," is the most used version. Good for decisions with real trade-offs, expensive in tokens.",[39,3190,3192],{"id":3191},"best-claude-skills-for-design-and-websites","Best Claude skills for design and websites",[11,3194,3195],{},"Three skills target the same problem, AI-generated interfaces that all look alike, from different angles.",[11,3197,3198,3048,3200,3204],{},[127,3199,2637],{},[18,3201,3203],{"href":2635,"rel":3202},[35],"nextlevelbuilder\u002Fui-ux-pro-max-skill",") is the broadest: a searchable library of styles, colour palettes, font pairings and UX rules that the agent consults while it builds. It is the one to try for a whole website or app.",[47,3206,3208],{"className":49,"code":3207,"language":51,"meta":52,"style":52},"\u002Fplugin marketplace add nextlevelbuilder\u002Fui-ux-pro-max-skill\n\u002Fplugin install ui-ux-pro-max@ui-ux-pro-max-skill\n",[54,3209,3210,3221],{"__ignoreMap":52},[57,3211,3212,3214,3216,3218],{"class":59,"line":60},[57,3213,64],{"class":63},[57,3215,86],{"class":67},[57,3217,89],{"class":67},[57,3219,3220],{"class":67}," nextlevelbuilder\u002Fui-ux-pro-max-skill\n",[57,3222,3223,3225,3227],{"class":59,"line":95},[57,3224,64],{"class":63},[57,3226,68],{"class":67},[57,3228,3229],{"class":67}," ui-ux-pro-max@ui-ux-pro-max-skill\n",[11,3231,3232,3048,3234,3238],{},[127,3233,2679],{},[18,3235,3237],{"href":2677,"rel":3236},[35],"Leonxlnx\u002Ftaste-skill",") is a family of variants (minimalist, brutalist, soft, redesign, image-to-code and more), so you pick a direction rather than a generic \"make it good\".",[47,3240,3242],{"className":49,"code":3241,"language":51,"meta":52,"style":52},"npx skills add https:\u002F\u002Fgithub.com\u002FLeonxlnx\u002Ftaste-skill\n",[54,3243,3244],{"__ignoreMap":52},[57,3245,3246,3248,3250,3252],{"class":59,"line":60},[57,3247,1223],{"class":63},[57,3249,1226],{"class":67},[57,3251,89],{"class":67},[57,3253,3254],{"class":67}," https:\u002F\u002Fgithub.com\u002FLeonxlnx\u002Ftaste-skill\n",[11,3256,3257,3259,3260,3048,3263,3268],{},[127,3258,2935],{}," (official, above) is the conservative choice and already available in Claude.ai. ",[127,3261,3262],{},"hallmark",[18,3264,3267],{"href":3265,"rel":3266},"https:\u002F\u002Fgithub.com\u002FNutlope\u002Fhallmark",[35],"Nutlope\u002Fhallmark",") is a newer anti-slop design skill that works across Claude Code, Cursor and Codex.",[11,3270,3271],{},"As with writing skills, use one design skill per project. They each have opinions about type and colour, and two sets of opinions average out to the generic look you were trying to avoid.",[11,3273,3274,3048,3277,3282],{},[127,3275,3276],{},"banana-claude",[18,3278,3281],{"href":3279,"rel":3280},"https:\u002F\u002Fgithub.com\u002FAgriciDaniel\u002Fbanana-claude",[35],"AgriciDaniel\u002Fbanana-claude",") adds image generation through Google's Gemini image model (\"Nano Banana\") to Claude Code. It installs disabled, because each image is a paid API call; enable it deliberately.",[39,3284,3286],{"id":3285},"best-claude-skills-for-writing","Best Claude skills for writing",[11,3288,3289],{},"The two most popular writing skills do the same job from different angles: they make Claude's prose sound less like Claude.",[11,3291,3292,3048,3294,3298],{},[127,3293,2742],{},[18,3295,3297],{"href":2740,"rel":3296},[35],"blader\u002Fhumanizer",") rewrites text to remove common AI patterns: \"not X but Y\" constructions, forced lists of three, dashes everywhere, inflated vocabulary.",[47,3300,3302],{"className":49,"code":3301,"language":51,"meta":52,"style":52},"\u002Fplugin marketplace add blader\u002Fhumanizer\n\u002Fplugin install humanizer@humanizer\n",[54,3303,3304,3315],{"__ignoreMap":52},[57,3305,3306,3308,3310,3312],{"class":59,"line":60},[57,3307,64],{"class":63},[57,3309,86],{"class":67},[57,3311,89],{"class":67},[57,3313,3314],{"class":67}," blader\u002Fhumanizer\n",[57,3316,3317,3319,3321],{"class":59,"line":95},[57,3318,64],{"class":63},[57,3320,68],{"class":67},[57,3322,3323],{"class":67}," humanizer@humanizer\n",[11,3325,3326,3048,3329,3334,3335,3337],{},[127,3327,3328],{},"stop-slop",[18,3330,3333],{"href":3331,"rel":3332},"https:\u002F\u002Fgithub.com\u002Fhardikpandya\u002Fstop-slop",[35],"hardikpandya\u002Fstop-slop",") is a skill folder with reference files of phrases, structures and examples to avoid. There is no install command; add the folder to ",[54,3336,2118],{},", or upload it in Claude.ai.",[11,3339,3340],{},"Use one, not both: they overlap, and two sets of style rules in context make the model second-guess itself.",[39,3342,3344],{"id":3343},"best-claude-skills-for-research-and-knowledge-work","Best Claude skills for research and knowledge work",[11,3346,3347,3048,3349,3353],{},[127,3348,2721],{},[18,3350,3352],{"href":2719,"rel":3351},[35],"mvanhorn\u002Flast30days-skill",") searches Reddit, Hacker News, GitHub and Polymarket out of the box, and X, YouTube and more once you add keys, then scores what it finds by engagement and writes one brief. Good for \"what are people saying about X this month\"; less good where the answer is in documentation rather than discussion.",[47,3355,3357],{"className":49,"code":3356,"language":51,"meta":52,"style":52},"\u002Fplugin marketplace add mvanhorn\u002Flast30days-skill\n\u002Fplugin install last30days\n",[54,3358,3359,3370],{"__ignoreMap":52},[57,3360,3361,3363,3365,3367],{"class":59,"line":60},[57,3362,64],{"class":63},[57,3364,86],{"class":67},[57,3366,89],{"class":67},[57,3368,3369],{"class":67}," mvanhorn\u002Flast30days-skill\n",[57,3371,3372,3374,3376],{"class":59,"line":95},[57,3373,64],{"class":63},[57,3375,68],{"class":67},[57,3377,3378],{"class":67}," last30days\n",[11,3380,3381,3048,3383,3387],{},[127,3382,2784],{},[18,3384,3386],{"href":2782,"rel":3385},[35],"kepano\u002Fobsidian-skills","), from Obsidian's CEO, teaches the agent Obsidian's CLI and its file formats (Markdown, Bases, Canvas), so Claude can work directly in a vault.",[47,3389,3391],{"className":49,"code":3390,"language":51,"meta":52,"style":52},"\u002Fplugin marketplace add kepano\u002Fobsidian-skills\n\u002Fplugin install obsidian@obsidian-skills\n",[54,3392,3393,3404],{"__ignoreMap":52},[57,3394,3395,3397,3399,3401],{"class":59,"line":60},[57,3396,64],{"class":63},[57,3398,86],{"class":67},[57,3400,89],{"class":67},[57,3402,3403],{"class":67}," kepano\u002Fobsidian-skills\n",[57,3405,3406,3408,3410],{"class":59,"line":95},[57,3407,64],{"class":63},[57,3409,68],{"class":67},[57,3411,3412],{"class":67}," obsidian@obsidian-skills\n",[11,3414,3415,3048,3417,3421,3422,3427,3428,3433],{},[127,3416,2700],{},[18,3418,3420],{"href":2698,"rel":3419},[35],"tt-a1i\u002Farchify",") turns an idea, plan or codebase into an interactive architecture diagram. For academic work, ",[18,3423,3426],{"href":3424,"rel":3425},"https:\u002F\u002Fgithub.com\u002FImbad0202\u002Facademic-research-skills",[35],"academic-research-skills"," covers the research, write, review and revise loop, and ",[18,3429,3432],{"href":3430,"rel":3431},"https:\u002F\u002Fgithub.com\u002FK-Dense-AI\u002Fscientific-agent-skills",[35],"K-Dense-AI\u002Fscientific-agent-skills"," is a large library for scientific workflows.",[39,3435,3437],{"id":3436},"best-claude-skills-for-marketing","Best Claude skills for marketing",[11,3439,3440,3441,3048,3443,3447],{},"The largest collection is ",[127,3442,2763],{},[18,3444,3446],{"href":2761,"rel":3445},[35],"coreyhaines31\u002Fmarketingskills","), MIT-licensed, with around fifty skills: copywriting, CRO, SEO audit, programmatic SEO, AI SEO, pricing, launch, cold email, ads, analytics and more.",[47,3449,3451],{"className":49,"code":3450,"language":51,"meta":52,"style":52},"npx skills add coreyhaines31\u002Fmarketingskills\n",[54,3452,3453],{"__ignoreMap":52},[57,3454,3455,3457,3459,3461],{"class":59,"line":60},[57,3456,1223],{"class":63},[57,3458,1226],{"class":67},[57,3460,89],{"class":67},[57,3462,3463],{"class":67}," coreyhaines31\u002Fmarketingskills\n",[11,3465,3466,3467,3469,3470,3474],{},"We cover it, the alternatives and how to combine them with live data from Google Ads, GA4 and Search Console in ",[18,3468,1414],{"href":1413},". A marketing skill knows how to write an ad; it does not know how your ads performed last week. That part comes from an MCP server, and our ",[18,3471,3473],{"href":3472},"\u002Fen\u002Fmcp-directory","MCP directory"," lists which marketing tools have one.",[39,3476,3478],{"id":3477},"skills-plugins-and-marketplaces","Skills, plugins and marketplaces",[11,3480,3481],{},"Searches for \"best Claude skills and plugins\" mix three things that are worth keeping apart:",[342,3483,3484,3493,3499],{},[124,3485,3486,3489,3490,3492],{},[127,3487,3488],{},"A skill"," is one folder with a ",[54,3491,482],{}," file. Claude reads its description, and loads the rest when a task matches.",[124,3494,3495,3498],{},[127,3496,3497],{},"A plugin"," is a Claude Code package that can bundle several skills with slash commands, subagents, hooks and MCP servers. superpowers and marketingskills are plugins that ship many skills.",[124,3500,3501,3504,3505,3508,3509,3512,3513,3518],{},[127,3502,3503],{},"A marketplace"," is a catalogue of plugins, usually a GitHub repository. ",[54,3506,3507],{},"\u002Fplugin marketplace add owner\u002Frepo"," adds one; ",[54,3510,3511],{},"\u002Fplugin install name@marketplace"," installs from it. ",[18,3514,3517],{"href":3515,"rel":3516},"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fclaude-plugins-official",[35],"anthropics\u002Fclaude-plugins-official"," is Anthropic's own.",[11,3520,3521,3522,3048,3525,3530],{},"Claude.ai has skills but no plugins, so there you upload the skill folder itself. ",[54,3523,3524],{},"npx skills",[18,3526,3529],{"href":3527,"rel":3528},"https:\u002F\u002Fgithub.com\u002Fvercel-labs\u002Fskills",[35],"vercel-labs\u002Fskills",") sits beside both: it installs the same skill folders into Claude Code, Cursor, Codex and other agents.",[39,3532,3534],{"id":3533},"check-a-skill-before-you-enable-it","Check a skill before you enable it",[11,3536,3537],{},"Popularity is not a security review. NVIDIA analysed over 31,000 public skills and found vulnerabilities in 26.1% and likely malicious intent in 5.2%: instructions that exfiltrate data, scripts that fetch and run remote code, prompt injection hidden in reference files.",[11,3539,3540],{},"Two steps cover most of the risk:",[121,3542,3543,3555],{},[124,3544,3545,1276,3548,3550,3551,3554],{},[127,3546,3547],{},"Read it.",[54,3549,482],{}," in full and every script. Look for anything that sends data out, touches credentials, writes outside the project or tells the agent to skip confirmations. The ",[18,3552,3553],{"href":2547},"five-minute checklist"," lists what to look for.",[124,3556,3557,1276,3560,3565,3566,27],{},[127,3558,3559],{},"Scan it.",[18,3561,3564],{"href":3562,"rel":3563},"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002FSkillSpector",[35],"NVIDIA SkillSpector"," is an open-source scanner for agent skills: 71 patterns across prompt injection, data exfiltration, privilege escalation, supply chain and MCP tool poisoning, with a 0–100 risk score. It reads a Git repository, a zip or a folder, so you can scan before anything lands in ",[54,3567,2118],{},[11,3569,3570],{},"Then pin the version you reviewed. A skill installed from a moving branch can change under you.",[39,3572,3574],{"id":3573},"where-to-find-more","Where to find more",[342,3576,3577,3585,3596,3618],{},[124,3578,3579,3584],{},[127,3580,3581],{},[18,3582,3517],{"href":3515,"rel":3583},[35],": the official Claude Code plugin directory, including skill-creator, frontend-design and superpowers.",[124,3586,3587,3048,3592,3595],{},[127,3588,3589,3590],{},"skills.sh and ",[54,3591,3524],{},[18,3593,3529],{"href":3527,"rel":3594},[35],"): a cross-agent installer and directory.",[124,3597,3598,1031,3601,1239,3606,3611,3612,3617],{},[127,3599,3600],{},"Curated lists",[18,3602,3605],{"href":3603,"rel":3604},"https:\u002F\u002Fgithub.com\u002FComposioHQ\u002Fawesome-claude-skills",[35],"ComposioHQ\u002Fawesome-claude-skills",[18,3607,3610],{"href":3608,"rel":3609},"https:\u002F\u002Fgithub.com\u002FVoltAgent\u002Fawesome-agent-skills",[35],"VoltAgent\u002Fawesome-agent-skills"," (1,000+ skills, including official ones from vendors' own teams) and ",[18,3613,3616],{"href":3614,"rel":3615},"https:\u002F\u002Fgithub.com\u002Fhesreallyhim\u002Fawesome-claude-code",[35],"hesreallyhim\u002Fawesome-claude-code",". Useful for discovery, but they list without reviewing.",[124,3619,3620,3627],{},[127,3621,3622],{},[18,3623,3626],{"href":3624,"rel":3625},"https:\u002F\u002Fgithub.com\u002FNVIDIA\u002Fskills",[35],"NVIDIA\u002Fskills",": a smaller catalogue of skills that passed NVIDIA's scanning and signing pipeline.",[39,3629,3631],{"id":3630},"how-to-choose-in-practice","How to choose, in practice",[121,3633,3634,3640,3646,3652],{},[124,3635,3636,3639],{},[127,3637,3638],{},"Start from a task you do every week",", not from a list. A skill earns its place by changing a recurring result.",[124,3641,3642,3645],{},[127,3643,3644],{},"Install two or three, not twenty."," Every skill's description sits in the agent's context, and overlapping skills compete.",[124,3647,3648,3651],{},[127,3649,3650],{},"Review and scan before installing",", pin a version, and keep a team copy of the ones you adopt.",[124,3653,3654,3657,3658,3660],{},[127,3655,3656],{},"Write your own for what is specific to you."," The best skill for \"how we write release notes\" is the one you write. ",[18,3659,1419],{"href":1418}," makes that a half-hour job.",[39,3662,1463],{"id":1462},[11,3664,3665],{},"Everything above is easy for one developer and hard for fifty. Once a team uses skills, three questions come up:",[342,3667,3668,3674,3680],{},[124,3669,3670,3673],{},[127,3671,3672],{},"Which skills are approved?"," Someone has to review and scan them once, so that every employee does not have to.",[124,3675,3676,3679],{},[127,3677,3678],{},"Is everyone on the same version?"," A team skill for \"how we review pull requests\" only helps if every agent runs the same copy, updated centrally.",[124,3681,3682,3685],{},[127,3683,3684],{},"Who may run what?"," A skill that deploys or sends email should only work for people allowed to do that. Skills cannot enforce it; the tools they call can.",[11,3687,3688,3689,3692],{},"That is the same governance question as which MCP servers a company allows, and it has the same answer: one approved catalogue, one policy, one log. Walma AI Hub distributes approved skills and MCP servers to every employee's Claude client, decides per role which tools an agent may call, and logs every call, inside the customer's own EU region. ",[18,3690,3691],{"href":424},"See how AI Hub works",", or book a 20-minute walkthrough below.",[427,3694,3695],{},"html pre.shiki code .svObZ, html code.shiki .svObZ{--shiki-default:#B392F0}html pre.shiki code .sU2Wk, html code.shiki .sU2Wk{--shiki-default:#9ECBFF}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 pre.shiki code .sDLfK, html code.shiki .sDLfK{--shiki-default:#79B8FF}",{"title":52,"searchDepth":95,"depth":431,"links":3697},[3698,3699,3700,3701,3702,3703,3704,3705,3706,3707,3708,3709,3710,3711],{"id":2560,"depth":95,"text":2561},{"id":2799,"depth":95,"text":2800},{"id":2846,"depth":95,"text":2847},{"id":3042,"depth":95,"text":3043},{"id":3139,"depth":95,"text":3140},{"id":3191,"depth":95,"text":3192},{"id":3285,"depth":95,"text":3286},{"id":3343,"depth":95,"text":3344},{"id":3436,"depth":95,"text":3437},{"id":3477,"depth":95,"text":3478},{"id":3533,"depth":95,"text":3534},{"id":3573,"depth":95,"text":3574},{"id":3630,"depth":95,"text":3631},{"id":1462,"depth":95,"text":1463},"2026-10-06","The top Claude skills for Claude Code and Claude.ai, ranked and grouped by job: coding, planning, design and websites, writing, research, marketing and documents. With GitHub stars, install commands, and how to check a skill before you enable it.",[3715,3718,3721,3724,3727,3730,3733,3736,3739,3742,3745],{"q":3716,"a":3717},"What are the best Claude skills?","It depends on the work. For documents, Anthropic's official docx, xlsx, pptx and pdf skills. For software teams, obra\u002Fsuperpowers for planning, test-driven development and debugging. For long tasks, planning-with-files. For UI and websites, frontend-design, taste-skill or ui-ux-pro-max. For writing, humanizer or stop-slop. For research, last30days. For marketing, coreyhaines31\u002Fmarketingskills. For building your own, skill-creator. Start with two or three that match a task you do every week.",{"q":3719,"a":3720},"What are the most popular Claude skills?","By GitHub stars on 12 October 2026: obra\u002Fsuperpowers (about 298,000), anthropics\u002Fskills (about 181,000), ui-ux-pro-max-skill (about 135,000), caveman (about 111,000) and taste-skill (about 95,000). Stars measure attention, not quality, so read a skill before you install it however popular it is.",{"q":3722,"a":3723},"What are the best Claude skills for coding?","obra\u002Fsuperpowers is the most used: it makes the agent brainstorm, plan, write tests first and debug systematically. Add planning-with-files for long tasks that outlive the context window, caveman if token use or verbose answers bother you, webapp-testing from Anthropic for front-end testing, and mcp-builder if you connect internal systems. Pick one methodology collection, not several, because they give competing instructions.",{"q":3725,"a":3726},"What is the difference between a Claude skill and a plugin?","A skill is one folder with a SKILL.md file that teaches Claude a task. A plugin is a package for Claude Code that can bundle several skills together with slash commands, subagents, hooks and MCP servers. Plugins are distributed through marketplaces, which is why many skills are installed with \u002Fplugin install. Claude.ai has no plugins; there you upload the skill folder itself.",{"q":3728,"a":3729},"Where can I find Claude skills?","The official ones are in Anthropic's GitHub repository anthropics\u002Fskills and in the official Claude Code plugin marketplace, anthropics\u002Fclaude-plugins-official. Community skills live in individual GitHub repositories, in curated lists such as ComposioHQ\u002Fawesome-claude-skills and VoltAgent\u002Fawesome-agent-skills, and in the skills.sh directory used by the npx skills command.",{"q":3731,"a":3732},"How do I install a Claude skill?","In Claude Code, add a plugin marketplace with \u002Fplugin marketplace add owner\u002Frepo and install with \u002Fplugin install name@marketplace, or use npx skills add owner\u002Frepo, or copy the skill folder into ~\u002F.claude\u002Fskills\u002F or .claude\u002Fskills\u002F in a project. In Claude.ai, upload the skill folder as a ZIP under the Skills settings; code execution must be enabled.",{"q":3734,"a":3735},"How many Claude skills should I install?","Two or three to start, and rarely more than ten. Every installed skill puts its name and description into the agent's context, so a long list costs tokens on every request and makes it harder for Claude to pick the right one. Overlapping skills, such as two writing-style skills, also give conflicting instructions.",{"q":3737,"a":3738},"Do Claude Code skills work in Claude.ai?","The format is the same, so most do. Zip the skill folder and upload it under the Skills settings in Claude.ai, with code execution enabled. Skills that depend on Claude Code features, such as hooks, local command-line tools or a plugin's slash commands, will not behave the same way in the browser.",{"q":3740,"a":3741},"Are Claude skills free?","Most are. Anthropic's example skills are Apache 2.0, while its document skills (docx, pdf, pptx, xlsx) are source-available under a proprietary licence but usable in Claude. Most community skills are open source on GitHub. You still pay for the model tokens a skill uses, and some skills call paid services, such as image generation.",{"q":3743,"a":3744},"Are community Claude skills safe?","Not automatically. NVIDIA's analysis of over 31,000 public skills found vulnerabilities in 26.1% and likely malicious intent in 5.2%. Read SKILL.md and every script before installing, run a scanner such as NVIDIA's SkillSpector, pin a version, and prefer a reviewed team copy.",{"q":3746,"a":3747},"What should I check before enabling a community plugin or skill?","Read SKILL.md in full and look for instructions that send data anywhere or tell the agent to skip confirmations. Read every script for network calls and writes outside the project. Check which tools it pre-approves in allowed-tools. Then pin a specific commit or release, so the version you reviewed is the version that runs.",{},"\u002Fguides\u002Fskills\u002Fbest-claude-skills",{"title":2532,"description":3713},"guides\u002Fskills\u002Fbest-claude-skills","vo5L3SD2C93mr4FDzOzzyAZuWQlsq67AUNnVaAaku14",{"id":3754,"title":3755,"author":6,"body":3756,"date":3712,"description":4168,"extension":441,"faq":4169,"meta":4185,"navigation":462,"order":581,"path":4186,"readTime":4187,"seo":4188,"stem":4189,"topic":468,"translationId":469,"updated":3712,"__hash__":4190},"guides\u002Fguides\u002Fskills\u002Fclaude-skill-creator.md","Claude skill creator: build and test your own skills with skill-creator",{"type":8,"value":3757,"toc":4159},[3758,3764,3770,3776,3780,3785,3799,3802,3825,3830,3839,3843,3846,3909,3916,3920,3926,3959,3965,3969,4041,4044,4048,4054,4081,4084,4088,4091,4113,4119,4123,4129,4148,4157],[11,3759,3760,3761,3763],{},"Writing a skill is easy. Writing one that reliably changes Claude's results is not, and most people never find out which they have, because they never test it. ",[127,3762,1419],{}," is Anthropic's answer: a skill for building skills, which interviews you, writes the files, and then measures whether the skill actually does better than Claude without it.",[11,3765,3766,3767,27],{},"This guide covers how to install it, the workflow step by step, what the evaluation does, and how to get a skill from your laptop to your team. For what skills are in the first place, see ",[18,3768,3769],{"href":20},"Claude skills explained",[11,3771,3772,3773,3775],{},"Facts are from skill-creator's own ",[54,3774,482],{}," and the anthropics\u002Fskills repository as of October 2026.",[39,3777,3779],{"id":3778},"install","Install",[11,3781,3782,3784],{},[127,3783,1094],{},", on its own from the official marketplace:",[47,3786,3788],{"className":49,"code":3787,"language":51,"meta":52,"style":52},"\u002Fplugin install skill-creator@claude-plugins-official\n",[54,3789,3790],{"__ignoreMap":52},[57,3791,3792,3794,3796],{"class":59,"line":60},[57,3793,64],{"class":63},[57,3795,68],{"class":67},[57,3797,3798],{"class":67}," skill-creator@claude-plugins-official\n",[11,3800,3801],{},"Or as part of Anthropic's example skills:",[47,3803,3805],{"className":49,"code":3804,"language":51,"meta":52,"style":52},"\u002Fplugin marketplace add anthropics\u002Fskills\n\u002Fplugin install example-skills@anthropic-agent-skills\n",[54,3806,3807,3817],{"__ignoreMap":52},[57,3808,3809,3811,3813,3815],{"class":59,"line":60},[57,3810,64],{"class":63},[57,3812,86],{"class":67},[57,3814,89],{"class":67},[57,3816,2139],{"class":67},[57,3818,3819,3821,3823],{"class":59,"line":95},[57,3820,64],{"class":63},[57,3822,68],{"class":67},[57,3824,2157],{"class":67},[11,3826,3827,3829],{},[127,3828,1107],{},": Anthropic's example skills, skill-creator included, are available in the app. Enable it under the Skills settings; code execution must be on.",[11,3831,3832,3833,148,3836,27],{},"Then ask for what you want: ",[1236,3834,3835],{},"\"Create a skill for writing our release notes\"",[1236,3837,3838],{},"\"Improve the skill in .claude\u002Fskills\u002Fincident-review and test it\"",[39,3840,3842],{"id":3841},"the-workflow","The workflow",[11,3844,3845],{},"skill-creator follows a fixed loop, which is most of its value:",[121,3847,3848,3854,3862,3868,3874,3880,3886,3892,3898,3904],{},[124,3849,3850,3853],{},[127,3851,3852],{},"Capture intent."," It interviews you: what the task is, what good output looks like, what goes wrong today. It may research the domain.",[124,3855,3856,1433,3859,3861],{},[127,3857,3858],{},"Write the skill.",[54,3860,482],{}," following Anthropic's writing guidance: progressive disclosure (short instructions, details in reference files) and no surprises.",[124,3863,3864,3867],{},[127,3865,3866],{},"Write test prompts."," Realistic requests the skill should handle.",[124,3869,3870,3873],{},[127,3871,3872],{},"Run with and without the skill."," Each test runs in parallel sub-agents, once with the new skill and once without (or against the previous version).",[124,3875,3876,3879],{},[127,3877,3878],{},"Draft assertions"," while the runs execute: what must be true of a good answer.",[124,3881,3882,3885],{},[127,3883,3884],{},"Grade."," A grader agent scores each output against the assertions.",[124,3887,3888,3891],{},[127,3889,3890],{},"Benchmark."," Pass rate, time and tokens are aggregated as mean ± standard deviation, with the difference from the baseline.",[124,3893,3894,3897],{},[127,3895,3896],{},"Analyse."," An analyst agent looks for patterns in what failed.",[124,3899,3900,3903],{},[127,3901,3902],{},"Review."," An HTML viewer shows the results so you can give feedback.",[124,3905,3906],{},[127,3907,3908],{},"Revise and repeat.",[11,3910,3911,3912,3915],{},"Steps 4 to 7 are what almost nobody does by hand. They answer the question that matters: ",[1236,3913,3914],{},"is Claude better with this skill than without it?"," A surprising number of skills make no difference, and some make results worse by adding rules the model already followed.",[39,3917,3919],{"id":3918},"fixing-skills-that-never-trigger","Fixing skills that never trigger",[11,3921,3922,3923,3925],{},"The most common skill problem is not the instructions, it is the ",[127,3924,704],{},": Claude never loads the skill because the description does not match how people phrase requests. skill-creator has a dedicated optimisation step for this, in Claude Code:",[342,3927,3928,3935,3942,3952],{},[124,3929,3930,3931,3934],{},"It writes about ",[127,3932,3933],{},"20 queries",", some that should trigger the skill and some that should not.",[124,3936,3937,3938,3941],{},"It splits them ",[127,3939,3940],{},"60\u002F40"," into training and test sets.",[124,3943,3944,3945,3948,3949,27],{},"It runs each query ",[127,3946,3947],{},"three times"," and iterates on the description ",[127,3950,3951],{},"up to five times",[124,3953,3954,3955,3958],{},"It keeps the description that scores best on the ",[127,3956,3957],{},"test"," set, so it is not just memorising the training queries.",[11,3960,3961,3962,3964],{},"If a skill you already have is ignored, run this before rewriting anything else. Our ",[18,3963,2842],{"href":2491}," has the manual checklist.",[39,3966,3968],{"id":3967},"claude-code-vs-claudeai","Claude Code vs Claude.ai",[179,3970,3971,3981],{},[182,3972,3973],{},[185,3974,3975,3977,3979],{},[188,3976],{},[188,3978,1094],{},[188,3980,1107],{},[195,3982,3983,3993,4004,4014,4028],{},[185,3984,3985,3988,3991],{},[200,3986,3987],{},"Interview and write the skill",[200,3989,3990],{},"Yes",[200,3992,3990],{},[185,3994,3995,3998,4001],{},[200,3996,3997],{},"Test prompts",[200,3999,4000],{},"In parallel sub-agents",[200,4002,4003],{},"One at a time",[185,4005,4006,4009,4011],{},[200,4007,4008],{},"Baseline comparison and benchmark",[200,4010,3990],{},[200,4012,4013],{},"No",[185,4015,4016,4019,4026],{},[200,4017,4018],{},"Description optimisation",[200,4020,4021,4022,4025],{},"Yes (uses the ",[54,4023,4024],{},"claude"," CLI)",[200,4027,4013],{},[185,4029,4030,4037,4039],{},[200,4031,4032,4033,4036],{},"Package as a ",[54,4034,4035],{},".skill"," file",[200,4038,3990],{},[200,4040,3990],{},[11,4042,4043],{},"Use Claude.ai to draft a skill quickly; use Claude Code when you want to know it works. Claude for Mac also offers a separate way to create a skill by recording yourself doing a task in Cowork, on Pro, Max and Team plans.",[39,4045,4047],{"id":4046},"what-it-ships-with","What it ships with",[11,4049,4050,4051,4053],{},"skill-creator is itself a good example of a well-built skill: the instructions are in ",[54,4052,482],{},", and the deterministic work is in scripts.",[342,4055,4056,4069,4075],{},[124,4057,4058,4061,4062,4065,4066,4068],{},[127,4059,4060],{},"Scripts"," for running evals, aggregating benchmarks, generating reports, improving descriptions, validating a skill and packaging it (",[54,4063,4064],{},"package_skill.py"," builds a ",[54,4067,4035],{}," file).",[124,4070,4071,4074],{},[127,4072,4073],{},"Agents",": instructions for a grader, an analyser and a blind comparator.",[124,4076,4077,4080],{},[127,4078,4079],{},"An eval viewer"," that turns results into an HTML page.",[11,4082,4083],{},"It is Apache 2.0 licensed, so you can study and adapt it.",[39,4085,4087],{"id":4086},"a-worked-example","A worked example",[11,4089,4090],{},"Say your team writes a weekly customer-health summary from the CRM and support tool.",[121,4092,4093,4098,4101,4104,4107],{},[124,4094,4095],{},[1236,4096,4097],{},"\"Create a skill for our weekly customer health summary. It pulls accounts with open escalations from HubSpot and Zendesk and writes a one-page summary for the leadership meeting.\"",[124,4099,4100],{},"skill-creator asks what the leadership team cares about, what the format is, and what a bad summary looks like. Give it last month's best summary as the example.",[124,4102,4103],{},"It writes the skill and four or five test prompts (\"summary for this week\", \"only enterprise accounts\", \"an account with no escalations\").",[124,4105,4106],{},"The benchmark tells you how often each version follows the format and includes the right accounts, and what it costs in time and tokens. A typical outcome is that the skill version is far more consistent and somewhat more expensive per run; if it is not more consistent, the skill is not pulling its weight yet.",[124,4108,4109,4110,4112],{},"You adjust, rerun, and commit the folder to ",[54,4111,2122],{}," in the team repo.",[11,4114,4115,4116,4118],{},"The skill describes the procedure; the data comes from MCP servers. See ",[18,4117,490],{"href":489}," for how the two fit together.",[39,4120,4122],{"id":4121},"from-your-laptop-to-the-team","From your laptop to the team",[11,4124,4125,4126,4128],{},"A skill in your ",[54,4127,2118],{}," helps you. A skill the team uses needs three more things:",[342,4130,4131,4137,4142],{},[124,4132,4133,4136],{},[127,4134,4135],{},"A shared home."," Commit project skills to the repository, or publish them in an internal plugin marketplace.",[124,4138,4139,4141],{},[127,4140,3902],{}," Treat a skill like code: a pull request, a second pair of eyes on scripts and instructions.",[124,4143,4144,4147],{},[127,4145,4146],{},"Evals kept with the skill."," Keep the test prompts and assertions skill-creator wrote next to the skill, and rerun them when the skill or the model changes.",[11,4149,4150,4151,4153,4154,4156],{},"For a company, approved skills belong on the same list as approved MCP servers. Walma AI Hub distributes both to every employee's client under one policy, with every call logged, inside the customer's own EU region. ",[18,4152,425],{"href":424},", or browse ",[18,4155,26],{"href":25}," for ones worth starting from.",[427,4158,429],{},{"title":52,"searchDepth":95,"depth":431,"links":4160},[4161,4162,4163,4164,4165,4166,4167],{"id":3778,"depth":95,"text":3779},{"id":3841,"depth":95,"text":3842},{"id":3918,"depth":95,"text":3919},{"id":3967,"depth":95,"text":3968},{"id":4046,"depth":95,"text":4047},{"id":4086,"depth":95,"text":4087},{"id":4121,"depth":95,"text":4122},"skill-creator is Anthropic's official skill for building Claude skills. This guide covers how to install it in Claude Code and Claude.ai, its workflow from interview to evals, benchmarks against a baseline, description optimisation, and packaging a skill for your team.",[4170,4173,4176,4179,4182],{"q":4171,"a":4172},"What is the Claude skill creator?","skill-creator is an official skill from Anthropic, in the anthropics\u002Fskills repository, that helps Claude create new skills, improve existing ones and measure whether they actually work. It interviews you about the task, writes the SKILL.md, writes test prompts, runs them with and without the skill, grades the results and helps you iterate.",{"q":4174,"a":4175},"How do I install skill-creator in Claude Code?","Either install it on its own from the official marketplace with \u002Fplugin install skill-creator@claude-plugins-official, or add Anthropic's skills repository with \u002Fplugin marketplace add anthropics\u002Fskills and install the example-skills plugin, which includes it. Then ask Claude to create a skill for the task you have in mind.",{"q":4177,"a":4178},"Can I use skill-creator in Claude.ai?","Yes. Anthropic's example skills, including skill-creator, are available in Claude.ai. The same create-and-test loop works there, but tests run one at a time, there are no baseline comparisons or quantitative benchmarks, and description optimisation is not available because it needs the claude command-line tool. You can still package the result as a skill file.",{"q":4180,"a":4181},"Does skill-creator test the skills it makes?","Yes. In Claude Code it runs each test prompt with and without the skill in parallel sub-agents, grades the outputs against assertions, and aggregates pass rate, time and tokens into a benchmark with mean and standard deviation, plus the difference from the baseline. An HTML viewer lets you review the results and give feedback.",{"q":4183,"a":4184},"Can skill-creator improve a skill's description?","Yes, in Claude Code. It generates about 20 queries that should or should not trigger the skill, splits them into training and test sets, runs each several times, and iterates on the description up to five times, keeping the version that scores best on the test set. This is the fix for skills that never trigger.",{},"\u002Fguides\u002Fskills\u002Fclaude-skill-creator","8 min read",{"title":3755,"description":4168},"guides\u002Fskills\u002Fclaude-skill-creator","fX0k30S4Ghwxvqv0lzusm8kPnJ6Mp26SaFBTIiyJHvE",{"id":4192,"title":4193,"author":6,"body":4194,"date":3712,"description":4671,"extension":441,"faq":4672,"meta":4688,"navigation":462,"order":586,"path":4689,"readTime":4187,"seo":4690,"stem":4691,"topic":468,"translationId":469,"updated":3712,"__hash__":4692},"guides\u002Fguides\u002Fskills\u002Fclaude-marketing-skills.md","Claude marketing skills: the best collections and how to use them with real data",{"type":8,"value":4195,"toc":4664},[4196,4203,4208,4211,4215,4221,4299,4302,4336,4339,4363,4369,4373,4453,4460,4464,4475,4575,4578,4598,4605,4609,4616,4619,4624,4628,4654,4661],[11,4197,4198,4199,4202],{},"Ask Claude for landing page copy without any context and you get competent, generic copy. Give it a skill that encodes how a good conversion copywriter works (the research questions, the structure, the objections to answer) and the result is noticeably better. Give it that skill ",[1236,4200,4201],{},"and"," your actual campaign and analytics data, and it stops giving advice and starts doing analysis.",[11,4204,4205,4206,27],{},"This guide covers the best marketing skill collections, how to install them, and how to combine them with live data. For how skills work in general, see ",[18,4207,3769],{"href":20},[11,4209,4210],{},"Repositories and install commands were checked on GitHub on 6 October 2026.",[39,4212,4214],{"id":4213},"the-main-collection-marketingskills","The main collection: marketingskills",[11,4216,4217,4220],{},[18,4218,3446],{"href":2761,"rel":4219},[35]," is the most widely used marketing skill collection by a distance. It is MIT-licensed and contains around fifty skills:",[179,4222,4223,4233],{},[182,4224,4225],{},[185,4226,4227,4230],{},[188,4228,4229],{},"Area",[188,4231,4232],{},"Skills (selection)",[195,4234,4235,4243,4251,4259,4267,4275,4283,4291],{},[185,4236,4237,4240],{},[200,4238,4239],{},"Copy and content",[200,4241,4242],{},"copywriting, copy-editing, content-strategy, emails, social, video",[185,4244,4245,4248],{},[200,4246,4247],{},"Conversion",[200,4249,4250],{},"cro, signup, onboarding, popups, paywalls, ab-testing, offers",[185,4252,4253,4256],{},[200,4254,4255],{},"SEO",[200,4257,4258],{},"seo-audit, programmatic-seo, ai-seo, schema, site-architecture",[185,4260,4261,4264],{},[200,4262,4263],{},"Paid",[200,4265,4266],{},"ads, ad-creative",[185,4268,4269,4272],{},[200,4270,4271],{},"Strategy",[200,4273,4274],{},"marketing-plan, product-marketing, pricing, launch, competitor-profiling, customer-research, marketing-psychology",[185,4276,4277,4280],{},[200,4278,4279],{},"Growth",[200,4281,4282],{},"referrals, lead-magnets, free-tools, cold-email, prospecting, churn-prevention",[185,4284,4285,4288],{},[200,4286,4287],{},"Measurement",[200,4289,4290],{},"analytics, attribution",[185,4292,4293,4296],{},[200,4294,4295],{},"Decision support",[200,4297,4298],{},"marketing-council (several viewpoints argue a decision)",[11,4300,4301],{},"Install all of it, or only what you use:",[47,4303,4305],{"className":49,"code":4304,"language":51,"meta":52,"style":52},"npx skills add coreyhaines31\u002Fmarketingskills\nnpx skills add coreyhaines31\u002Fmarketingskills --skill cro copywriting\n",[54,4306,4307,4317],{"__ignoreMap":52},[57,4308,4309,4311,4313,4315],{"class":59,"line":60},[57,4310,1223],{"class":63},[57,4312,1226],{"class":67},[57,4314,89],{"class":67},[57,4316,3463],{"class":67},[57,4318,4319,4321,4323,4325,4328,4330,4333],{"class":59,"line":95},[57,4320,1223],{"class":63},[57,4322,1226],{"class":67},[57,4324,89],{"class":67},[57,4326,4327],{"class":67}," coreyhaines31\u002Fmarketingskills",[57,4329,3170],{"class":635},[57,4331,4332],{"class":67}," cro",[57,4334,4335],{"class":67}," copywriting\n",[11,4337,4338],{},"Or as a Claude Code plugin:",[47,4340,4342],{"className":49,"code":4341,"language":51,"meta":52,"style":52},"\u002Fplugin marketplace add coreyhaines31\u002Fmarketingskills\n\u002Fplugin install marketing-skills\n",[54,4343,4344,4354],{"__ignoreMap":52},[57,4345,4346,4348,4350,4352],{"class":59,"line":60},[57,4347,64],{"class":63},[57,4349,86],{"class":67},[57,4351,89],{"class":67},[57,4353,3463],{"class":67},[57,4355,4356,4358,4360],{"class":59,"line":95},[57,4357,64],{"class":63},[57,4359,68],{"class":67},[57,4361,4362],{"class":67}," marketing-skills\n",[11,4364,4365,4368],{},[127,4366,4367],{},"Install a subset."," Fifty skills means fifty descriptions in Claude's context, and several overlap (copywriting and copy-editing, ads and ad-creative). Pick the five to ten that match work you do every month.",[39,4370,4372],{"id":4371},"alternatives","Alternatives",[179,4374,4375,4387],{},[182,4376,4377],{},[185,4378,4379,4382,4385],{},[188,4380,4381],{},"Collection",[188,4383,4384],{},"Focus",[188,4386,3779],{},[195,4388,4389,4404,4421,4436],{},[185,4390,4391,4398,4401],{},[200,4392,4393],{},[18,4394,4397],{"href":4395,"rel":4396},"https:\u002F\u002Fgithub.com\u002Fnowork-studio\u002Fnotfair-plugin",[35],"nowork-studio\u002Fnotfair-plugin",[200,4399,4400],{},"SEO, GEO, Google, Meta and LinkedIn ads: keyword research, Google Ads audit, local SEO",[200,4402,4403],{},"Claude Code plugin marketplace",[185,4405,4406,4413,4416],{},[200,4407,4408],{},[18,4409,4412],{"href":4410,"rel":4411},"https:\u002F\u002Fgithub.com\u002Firinabuht12-oss\u002Fmarketing-skills",[35],"irinabuht12-oss\u002Fmarketing-skills",[200,4414,4415],{},"Paid media and analysis: A\u002FB test analysis, ad copy variants, channel mix",[200,4417,4418,4419],{},"Copy into ",[54,4420,2118],{},[185,4422,4423,4430,4433],{},[200,4424,4425],{},[18,4426,4429],{"href":4427,"rel":4428},"https:\u002F\u002Fgithub.com\u002Fericosiu\u002Fai-marketing-skills",[35],"ericosiu\u002Fai-marketing-skills",[200,4431,4432],{},"Growth operations: SEO ops, content ops, outbound, sales pipeline",[200,4434,4435],{},"Git clone",[185,4437,4438,4445,4448],{},[200,4439,4440],{},[18,4441,4444],{"href":4442,"rel":4443},"https:\u002F\u002Fgithub.com\u002Faaron-he-zhu\u002Faaron-marketing-skills",[35],"aaron-he-zhu\u002Faaron-marketing-skills",[200,4446,4447],{},"A broad set of more than a hundred skills",[200,4449,4450,4451],{},"Plugin marketplace or ",[54,4452,3524],{},[11,4454,4455,4456,4459],{},"If you mainly run paid search and social, a paid-media collection fits better than the general one. For writing, add a style skill such as ",[18,4457,2742],{"href":4458},"\u002Fen\u002Fguides\u002Fskills\u002Fbest-claude-skills#writing-remove-the-ai-tells"," to strip the AI tells from final copy.",[39,4461,4463],{"id":4462},"skills-give-know-how-mcp-gives-the-numbers","Skills give know-how. MCP gives the numbers.",[11,4465,4466,4467,4470,4471,4474],{},"This is the part most marketing skill tutorials skip. A skill like ",[54,4468,4469],{},"seo-audit"," knows ",[1236,4472,4473],{},"how"," to audit a site. Without data, it audits what you paste in, or what it can fetch from a public page. The useful version reads your own sources:",[179,4476,4477,4489],{},[182,4478,4479],{},[185,4480,4481,4484,4486],{},[188,4482,4483],{},"Data",[188,4485,1574],{},[188,4487,4488],{},"Our guide",[195,4490,4491,4505,4519,4533,4547,4561],{},[185,4492,4493,4496,4499],{},[200,4494,4495],{},"Search queries, clicks, positions",[200,4497,4498],{},"Google Search Console",[200,4500,4501],{},[18,4502,4504],{"href":4503},"\u002Fen\u002Fguides\u002Fmcp\u002Fgoogle-search-console-mcp","GSC MCP",[185,4506,4507,4510,4513],{},[200,4508,4509],{},"Sessions, conversions, channels",[200,4511,4512],{},"Google Analytics 4",[200,4514,4515],{},[18,4516,4518],{"href":4517},"\u002Fen\u002Fguides\u002Fmcp\u002Fgoogle-analytics-mcp","GA4 MCP",[185,4520,4521,4524,4527],{},[200,4522,4523],{},"Campaigns, search terms, spend",[200,4525,4526],{},"Google Ads",[200,4528,4529],{},[18,4530,4532],{"href":4531},"\u002Fen\u002Fguides\u002Fmcp\u002Fgoogle-ads-mcp","Google Ads MCP",[185,4534,4535,4538,4541],{},[200,4536,4537],{},"Keywords, backlinks, competitors",[200,4539,4540],{},"Ahrefs",[200,4542,4543],{},[18,4544,4546],{"href":4545},"\u002Fen\u002Fguides\u002Fmcp\u002Fahrefs-mcp","Ahrefs MCP",[185,4548,4549,4552,4555],{},[200,4550,4551],{},"Contacts, deals, lifecycle",[200,4553,4554],{},"HubSpot",[200,4556,4557],{},[18,4558,4560],{"href":4559},"\u002Fen\u002Fguides\u002Fmcp\u002Fhubspot-mcp","HubSpot MCP",[185,4562,4563,4566,4569],{},[200,4564,4565],{},"Product usage",[200,4567,4568],{},"PostHog",[200,4570,4571],{},[18,4572,4574],{"href":4573},"\u002Fen\u002Fguides\u002Fmcp\u002Fposthog-mcp","PostHog MCP",[11,4576,4577],{},"With both in place, requests change character:",[342,4579,4580,4586,4592],{},[124,4581,4582,4585],{},[1236,4583,4584],{},"\"Run a CRO review of our pricing page\""," becomes a review that cites the page's actual conversion rate and drop-off from GA4.",[124,4587,4588,4591],{},[1236,4589,4590],{},"\"Find SEO quick wins\""," becomes a list of your pages ranking 8 to 20 in Search Console, with the queries and impressions behind each.",[124,4593,4594,4597],{},[1236,4595,4596],{},"\"Audit our Google Ads\""," becomes an audit of last month's search terms, with the ones that spent money without converting.",[11,4599,4600,4601,27],{},"A summary of which servers matter for which marketing job is in ",[18,4602,4604],{"href":4603},"\u002Fen\u002Fguides\u002Fmcp\u002Fbest-mcp-servers-for-marketing-and-seo","the best MCP servers for marketing and SEO",[39,4606,4608],{"id":4607},"write-your-own-for-the-recurring-work","Write your own for the recurring work",[11,4610,4611,4612,4615],{},"The collections cover general marketing craft. What they cannot know is ",[1236,4613,4614],{},"your"," plan, your targets and your reporting format. That is where your own skill pays off.",[11,4617,4618],{},"We run our own weekly SEO review this way: a skill that knows our SEO plan, our targets per quarter and where the progress log lives, pulls Search Console, Ahrefs and Google Ads through MCP, and writes the same brief every Monday: what changed, what is behind target, and the one or two actions for the week. The procedure lives in the skill, the data comes from MCP, and the history lives in a log file in the repository.",[11,4620,4621,4623],{},[18,4622,1419],{"href":1418}," is the fastest way to build a skill like that, and it tests it against a baseline so you know it helps.",[39,4625,4627],{"id":4626},"a-setup-that-works-for-a-marketing-team","A setup that works for a marketing team",[121,4629,4630,4636,4642,4648],{},[124,4631,4632,4635],{},[127,4633,4634],{},"Pick five to ten skills"," from one collection, matching recurring work: copywriting, CRO, SEO audit, ads, analytics.",[124,4637,4638,4641],{},[127,4639,4640],{},"Connect the data",": Search Console, GA4, the ad platforms and your CRM through MCP, read-only to start.",[124,4643,4644,4647],{},[127,4645,4646],{},"Write one skill of your own"," for the recurring report everyone already does by hand.",[124,4649,4650,4653],{},[127,4651,4652],{},"Keep it consistent",": the same skills and the same connections for the whole team, so two people asking the same question get the same answer.",[11,4655,4656,4657,27],{},"The last step is the hard one with per-person setups. Walma AI Hub distributes approved skills and MCP connections to the whole marketing team under one policy, with every call logged, inside the customer's own EU region. The marketing-specific version is on ",[18,4658,4660],{"href":4659},"\u002Fen\u002Fai-for-marketing","AI for marketing",[427,4662,4663],{},"html pre.shiki code .svObZ, html code.shiki .svObZ{--shiki-default:#B392F0}html pre.shiki code .sU2Wk, html code.shiki .sU2Wk{--shiki-default:#9ECBFF}html pre.shiki code .sDLfK, html code.shiki .sDLfK{--shiki-default:#79B8FF}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);}",{"title":52,"searchDepth":95,"depth":431,"links":4665},[4666,4667,4668,4669,4670],{"id":4213,"depth":95,"text":4214},{"id":4371,"depth":95,"text":4372},{"id":4462,"depth":95,"text":4463},{"id":4607,"depth":95,"text":4608},{"id":4626,"depth":95,"text":4627},"The best Claude skills for marketing: coreyhaines31\u002Fmarketingskills and the alternatives, what each covers (copywriting, CRO, SEO, ads, email, launch), how to install them in Claude Code and Claude.ai, and how to combine them with live data from Google Ads, GA4 and Search Console through MCP.",[4673,4676,4679,4682,4685],{"q":4674,"a":4675},"What are Claude marketing skills?","Marketing skills are Claude skills, folders with a SKILL.md and supporting files, that teach Claude a marketing task: writing landing page copy, auditing a page for conversion, planning a launch, running an SEO audit or structuring an A\u002FB test. Claude loads the relevant one when your request matches it.",{"q":4677,"a":4678},"What is the best Claude marketing skills collection?","The largest and most used is coreyhaines31\u002Fmarketingskills, an MIT-licensed collection of around fifty skills covering copywriting, CRO, SEO, programmatic SEO, AI SEO, pricing, launch, email, ads, analytics and more. Alternatives focus on paid media (notfair-plugin, irinabuht12-oss\u002Fmarketing-skills) or on growth operations (ericosiu\u002Fai-marketing-skills).",{"q":4680,"a":4681},"How do I install marketing skills in Claude?","In Claude Code, run npx skills add coreyhaines31\u002Fmarketingskills, or add it as a plugin marketplace with \u002Fplugin marketplace add coreyhaines31\u002Fmarketingskills and install marketing-skills. You can install only some skills with --skill. In Claude.ai, upload the individual skill folders you want as ZIP files under the Skills settings.",{"q":4683,"a":4684},"Can Claude marketing skills read my Google Ads or Analytics data?","Not on their own. A skill is know-how: how to audit a campaign or structure a report. To work on real numbers, Claude also needs access to the data, which comes from MCP servers for Google Ads, Google Analytics, Search Console, Ahrefs and similar tools. Skill plus MCP server is the combination that produces analysis instead of generic advice.",{"q":4686,"a":4687},"Do I need Claude Code for marketing skills?","No. Skills also work in Claude.ai, where you upload them in settings and code execution must be enabled. Claude Code is better for recurring work, such as a weekly report, because the skills, the data connections and the output files live in one project folder.",{},"\u002Fguides\u002Fskills\u002Fclaude-marketing-skills",{"title":4193,"description":4671},"guides\u002Fskills\u002Fclaude-marketing-skills","AYXmpMmCBXFWAaVaJsEKPIorWf6crZo9JPK7bBINUjg",{"id":4,"title":5,"author":6,"body":4694,"date":439,"description":440,"extension":441,"faq":4987,"meta":4994,"navigation":462,"order":463,"path":464,"readTime":465,"seo":4995,"stem":467,"topic":468,"translationId":469,"updated":439,"__hash__":470},{"type":8,"value":4695,"toc":4979},[4696,4698,4704,4709,4711,4713,4725,4727,4749,4751,4755,4757,4759,4791,4793,4795,4921,4923,4925,4943,4945,4947,4965,4967,4969,4971,4977],[11,4697,13],{},[11,4699,16,4700,22,4702,27],{},[18,4701,21],{"href":20},[18,4703,26],{"href":25},[11,4705,30,4706,37],{},[18,4707,36],{"href":33,"rel":4708},[35],[39,4710,42],{"id":41},[11,4712,45],{},[47,4714,4715],{"className":49,"code":50,"language":51,"meta":52,"style":52},[54,4716,4717],{"__ignoreMap":52},[57,4718,4719,4721,4723],{"class":59,"line":60},[57,4720,64],{"class":63},[57,4722,68],{"class":67},[57,4724,71],{"class":67},[11,4726,74],{},[47,4728,4729],{"className":49,"code":77,"language":51,"meta":52,"style":52},[54,4730,4731,4741],{"__ignoreMap":52},[57,4732,4733,4735,4737,4739],{"class":59,"line":60},[57,4734,64],{"class":63},[57,4736,86],{"class":67},[57,4738,89],{"class":67},[57,4740,92],{"class":67},[57,4742,4743,4745,4747],{"class":59,"line":95},[57,4744,64],{"class":63},[57,4746,68],{"class":67},[57,4748,102],{"class":67},[11,4750,105],{},[11,4752,108,4753,112],{},[54,4754,111],{},[39,4756,116],{"id":115},[11,4758,119],{},[121,4760,4761,4765,4769,4773,4779,4783,4787],{},[124,4762,4763,130],{},[127,4764,129],{},[124,4766,4767,136],{},[127,4768,135],{},[124,4770,4771,142],{},[127,4772,141],{},[124,4774,4775,148,4777,152],{},[127,4776,147],{},[127,4778,151],{},[124,4780,4781,158],{},[127,4782,157],{},[124,4784,4785,164],{},[127,4786,163],{},[124,4788,4789,170],{},[127,4790,169],{},[11,4792,173],{},[39,4794,177],{"id":176},[179,4796,4797,4805],{},[182,4798,4799],{},[185,4800,4801,4803],{},[188,4802,190],{},[188,4804,193],{},[195,4806,4807,4815,4823,4831,4839,4847,4855,4863,4871,4881,4889,4897,4905,4913],{},[185,4808,4809,4813],{},[200,4810,4811],{},[127,4812,129],{},[200,4814,206],{},[185,4816,4817,4821],{},[200,4818,4819],{},[127,4820,141],{},[200,4822,215],{},[185,4824,4825,4829],{},[200,4826,4827],{},[127,4828,151],{},[200,4830,224],{},[185,4832,4833,4837],{},[200,4834,4835],{},[127,4836,147],{},[200,4838,233],{},[185,4840,4841,4845],{},[200,4842,4843],{},[127,4844,240],{},[200,4846,243],{},[185,4848,4849,4853],{},[200,4850,4851],{},[127,4852,157],{},[200,4854,252],{},[185,4856,4857,4861],{},[200,4858,4859],{},[127,4860,259],{},[200,4862,262],{},[185,4864,4865,4869],{},[200,4866,4867],{},[127,4868,269],{},[200,4870,272],{},[185,4872,4873,4879],{},[200,4874,4875,279,4877],{},[127,4876,163],{},[127,4878,282],{},[200,4880,285],{},[185,4882,4883,4887],{},[200,4884,4885],{},[127,4886,135],{},[200,4888,294],{},[185,4890,4891,4895],{},[200,4892,4893],{},[127,4894,169],{},[200,4896,303],{},[185,4898,4899,4903],{},[200,4900,4901],{},[127,4902,310],{},[200,4904,313],{},[185,4906,4907,4911],{},[200,4908,4909],{},[127,4910,320],{},[200,4912,323],{},[185,4914,4915,4919],{},[200,4916,4917],{},[127,4918,330],{},[200,4920,333],{},[39,4922,337],{"id":336},[11,4924,340],{},[342,4926,4927,4931,4935],{},[124,4928,4929,349],{},[127,4930,348],{},[124,4932,4933,355],{},[127,4934,354],{},[124,4936,4937,361,4939,366,4941,27],{},[127,4938,360],{},[18,4940,365],{"href":364},[18,4942,370],{"href":369},[11,4944,373],{},[39,4946,377],{"id":376},[342,4948,4949,4953,4957,4961],{},[124,4950,4951,385],{},[127,4952,384],{},[124,4954,4955,391],{},[127,4956,390],{},[124,4958,4959,397],{},[127,4960,396],{},[124,4962,4963,403],{},[127,4964,402],{},[11,4966,406],{},[39,4968,410],{"id":409},[11,4970,413],{},[11,4972,416,4973,421,4975,27],{},[18,4974,420],{"href":419},[18,4976,425],{"href":424},[427,4978,429],{},{"title":52,"searchDepth":95,"depth":431,"links":4980},[4981,4982,4983,4984,4985,4986],{"id":41,"depth":95,"text":42},{"id":115,"depth":95,"text":116},{"id":176,"depth":95,"text":177},{"id":336,"depth":95,"text":337},{"id":376,"depth":95,"text":377},{"id":409,"depth":95,"text":410},[4988,4989,4990,4991,4992,4993],{"q":444,"a":445},{"q":447,"a":448},{"q":450,"a":451},{"q":453,"a":454},{"q":456,"a":457},{"q":459,"a":460},{},{"title":5,"description":440},[],1792405295656]