AI tools · Skill

The best agent skills

A skill is a set of instructions that teaches the AI a method: how to debug, how to scope a project, how to review code. It steers the agent’s behaviour, not what it can access.

28 scored skill. Last checked on 25 October 2026.

Photo & image 1 Video 1 Design & slides 7 Code 11 Office & email 1 Route & save 1 Feed (memory, context) 1 Frame (methods, guardrails) 20

The three to remember

Most powerful

Frontend Design (Anthropic)

SkillRisingSecurity: nothing to flag

9.6

The most installed skill on the official marketplace: one click, immediate effect on UI quality.

· Utility 9/10, best of the selection
· Free, no limits
· No known security caveat

Free1 clickSite ↗

Simplest

React best practices (Vercel)

SkillRisingSecurity: nothing to flag

9.2

The reference skill for clean React, maintained by the makers of Next.js.

· Utility 8/10
· Free, no limits
· No known security caveat

Free1 clickSite ↗

The alternative

grill-me (Matt Pocock)

SkillSecurity: nothing to flag

9.1

A small skill with a big impact: it catches gaps in a specification before they become costly. Also works outside code to challenge any project.

· Utility 8/10
· Free, no limits
· No known security caveat

Free1 clickSite ↗

Claude Code 20 Any agent 4 Cursor 2 other 2

For Claude Code (20)

Tool	Type	What it does	Price	Install	Score	Sec.
Frontend Design (Anthropic)RisingOfficial Anthropic skill that pushes Claude to produce polished, distinctive interfaces rather than generic templates.	Skill	Official Anthropic skill that pushes Claude to produce polished, distinctive interfaces rather than generic templates.	Free	1 click	9.6	Security: nothing to flag
React best practices (Vercel)RisingReact and Next.js rules maintained by Vercel, read by the agent before coding.	Skill	React and Next.js rules maintained by Vercel, read by the agent before coding.	Free	1 click	9.2	Security: nothing to flag
grill-me (Matt Pocock)The agent questions you relentlessly on your plan, branch by branch, until every decision is clear, before writing a single line of code.	Skill	The agent questions you relentlessly on your plan, branch by branch, until every decision is clear, before writing a single line of code.	Free	1 click	9.1	Security: nothing to flag
UI UX Pro MaxDesign skill that gives the agent a base of styles, palettes, font pairings, and UX rules before it writes the interface.	Skill	Design skill that gives the agent a base of styles, palettes, font pairings, and UX rules before it writes the interface.	Free	1 click	9.1	Security: nothing to flag
awesome-claude-codeRisingA curated list of commands, hooks, skills, and workflows for Claude Code.	Skill	A curated list of commands, hooks, skills, and workflows for Claude Code.	Free	1 click	9.0	Security: nothing to flag
CavemanMakes the agent respond in telegraphic style: fewer pleasantries and filler, code and error messages kept exactly as-is.	Skill	Makes the agent respond in telegraphic style: fewer pleasantries and filler, code and error messages kept exactly as-is.	Free	1 click	8.8	Security: nothing to flag
Questionnaire fournisseur (skill nAIvigate)Pre-fills a security questionnaire from your policies, tagging each answer Sourced, Inferred or To complete.	Skill	Pre-fills a security questionnaire from your policies, tagging each answer Sourced, Inferred or To complete.	Free	1 click	8.8	Security: nothing to flag
mcp-builder (Anthropic)RisingOfficial skill that helps design and code a clean MCP server.	Skill	Official skill that helps design and code a clean MCP server.	Free	1 click	8.7	Security: nothing to flag
Taste Skill (dont minimalist-ui)Collection of "anti-slop" skills for front-end work: bans purple gradients, Inter everywhere, and rows of rounded cards; variants include minimalist-ui, etc.	Skill	Collection of "anti-slop" skills for front-end work: bans purple gradients, Inter everywhere, and rows of rounded cards; variants include minimalist-ui, etc.	Free	1 click	8.7	Security: nothing to flag
skill-creator (Anthropic)RisingOfficial skill that guides Claude to write, structure, and test a new skill.	Skill	Official skill that guides Claude to write, structure, and test a new skill.	Free	1 click	8.7	Security: nothing to flag
brand-guidelines (Anthropic)RisingExample skill that applies a brand charter (colors, typography, tone) to what Claude produces.	Skill	Example skill that applies a brand charter (colors, typography, tone) to what Claude produces.	Free	1 click	8.4	Security: nothing to flag
Revue de PSSI (skill nAIvigate)Reviews a security policy and lists the gaps against ISO 27001:2022, with corrected wording.	Skill	Reviews a security policy and lists the gaps against ISO 27001:2022, with corrected wording.	Free	1 click	8.3	Security: nothing to flag
Cartographie AI Act (skill nAIvigate)Sorts your AI use cases by AI Act risk level, with role, obligations, deadlines and points to confirm.	Skill	Sorts your AI use cases by AI Act risk level, with role, obligations, deadlines and points to confirm.	Free	1 click	8.3	Security: nothing to flag
Veille CVE ciblée (skill nAIvigate)Filters vulnerability bulletins against your actual estate and sorts them into 72 h, 30 days, to plan.	Skill	Filters vulnerability bulletins against your actual estate and sorts them into 72 h, 30 days, to plan.	Free	1 click	8.3	Security: nothing to flag
canvas-design (Anthropic)Official skill that makes Claude produce polished static visuals (posters, visuals, compositions) in PNG or PDF.	Skill	Official skill that makes Claude produce polished static visuals (posters, visuals, compositions) in PNG or PDF.	Free	1 click	8.2	Security: nothing to flag
web-design-guidelines (Vercel)Audits your interface code against more than one hundred Vercel rules (accessibility, forms, animations, images) and reports gaps file by file.	Skill	Audits your interface code against more than one hundred Vercel rules (accessibility, forms, animations, images) and reports gaps file by file.	Free	1 click	8.1	Security: one caveat, see the card
Skills documents Anthropic (docx, xlsx, pptx, pdf)RisingThe four official skills for reading, creating and editing Word, Excel, PowerPoint and PDF files from Claude.	Skill	The four official skills for reading, creating and editing Word, Excel, PowerPoint and PDF files from Claude.	Free	1 click	8.0	Security: one caveat, see the card
Compte-rendu de comité sécurité (skill nAIvigate)Turns raw committee notes into decisions, dated actions with owners, open risks and deferred items.	Skill	Turns raw committee notes into decisions, dated actions with owners, open risks and deferred items.	Free	1 click	8.0	Security: nothing to flag
find-skills (Vercel)RisingSkill that searches and installs other skills from skills.sh on demand.	Skill	Skill that searches and installs other skills from skills.sh on demand.	Free	1 click	7.9	Security: one caveat, see the card
video-use (Browser Use)Video editing by agent: drop your footage in a folder, describe the result, the agent cuts silences, colour grades, adds subtitles and delivers final.mp4.	Skill	Video editing by agent: drop your footage in a folder, describe the result, the agent cuts silences, colour grades, adds subtitles and delivers final.mp4.	Free	5 min	7.5	Security: one caveat, see the card

For Any agent (4)

Tool	Type	What it does	Price	Install	Score	Sec.
BMAD MethodRisingAgile method for agents: roles (PM, architect, dev) and structured documents before the code.	Skill	Agile method for agents: roles (PM, architect, dev) and structured documents before the code.	Free	5 min	9.0	Security: nothing to flag
AGENTS.md (convention)A file of instructions at the root of the repository, read by most coding agents.	Skill	A file of instructions at the root of the repository, read by most coding agents.	Free	5 min	7.9	Security: nothing to flag
awesome-agent-skills (VoltAgent)RisingCatalogue of more than one thousand multi-agent skills, organised by domain.	Skill	Catalogue of more than one thousand multi-agent skills, organised by domain.	Free	1 click	7.9	Security: one caveat, see the card
andrej-karpathy-skillsRisingPortable skills inspired by Andrej Karpathy's practices for coding with an agent.	Skill	Portable skills inspired by Andrej Karpathy's practices for coding with an agent.	Free	5 min	7.6	Security: one caveat, see the card

For Cursor (2)

Tool	Type	What it does	Price	Install	Score	Sec.
Awesome CursorRulesCommunity library of rule files by framework and language to guide Cursor.	Skill	Community library of rule files by framework and language to guide Cursor.	Free	5 min	8.8	Security: nothing to flag
cursor.directoryRisingDirectory of Cursor rules and MCP servers, with prompts ready to paste.	Skill	Directory of Cursor rules and MCP servers, with prompts ready to paste.	Free	1 click	8.8	Security: nothing to flag

For other (2)

Tool	Type	What it does	Price	Install	Score	Sec.
Image to Editable PPT (Ming-Image)NewAgent skill that recreates an AI-generated slide image as an editable PowerPoint presentation, with texts and simple shapes converted to native elements.	Skill	Agent skill that recreates an AI-generated slide image as an editable PowerPoint presentation, with texts and simple shapes converted to native elements.	Free	Technical	6.8	Security: one caveat, see the card
Ling UI Design (Ming-Image)NewAn agent skill that generates a visual UI mockup, splits it into layers, then uses those layers to write UI code and visually verify the result.	Skill	An agent skill that generates a visual UI mockup, splits it into layers, then uses those layers to write UI code and visually verify the result.	Free	Technical	6.4	Security: one caveat, see the card

App : A product you use directly · Tool : A precise action the AI can call · MCP : Standardises the connection between an AI application and systems that expose tools or data · Built-in mode : A feature to switch on inside an assistant you already use (ChatGPT, Claude, Gemini): nothing to install · Skill : A reusable method: instructions, procedures, sometimes scripts or templates · Plugin : A set of capabilities installed into an AI application or a host

What to know

A skill is text, but text the agent follows. Read it before installing, exactly as you would read a script, and prefer official marketplaces to unknown repositories.

Find the right one in 4 questions →