What Are Claude Skills
Plain-English answer to what a Claude Skill is, plus 12 named skills people actually use, an annotated SKILL.md, and the 5-way matrix vs MCP, subagents, plugins, projects.
Arrêtez de configurer. Commencez à construire.
Templates SaaS avec orchestration IA.
Problem: You keep seeing "Claude Skills" in repos, on X, in YouTube thumbnails, and you still are not sure what the thing is. Anthropic's docs read like a spec. The blog posts in the search results bury the answer under five paragraphs of intro.
Quick Win: A Claude Skill is a small folder with one file inside it called SKILL.md. That file tells Claude how to do one specific thing. Build the folder once. Claude uses it forever.
That sentence is the whole concept. The rest of this post is the proof, the parts list, the 12 real ones people are running today, and the matrix that tells you when to reach for a skill instead of an MCP server, a subagent, a plugin, or a project.
A One-Paragraph Definition You Can Quote
A Claude Skill is a folder of instructions Claude reads when it decides the topic fits your request. The folder lives at .claude/skills/your-skill-name/ and contains a single required file, SKILL.md. Claude scans the descriptions of every skill you have installed. When one matches what you asked for, it loads the body and follows it. You can also call a skill by hand with /skill-name. Anthropic's own line is: "Skills are folders containing instructions, scripts, and resources that Claude discovers and loads dynamically when relevant to a task." (Source: claude.com/blog/skills-explained.)
Three things follow from that definition. Skills sit on disk. Skills are discovered by description-matching, not by file extension or folder location. Skills are reusable across every session in the project they live in.
A Real SKILL.md File, Opened Up
Here is a working skill in ten lines, taken from Anthropic's own docs at code.claude.com/docs/en/skills:
---
description: Summarizes uncommitted changes and flags anything risky. Use when the user asks what changed, wants a commit message, or asks to review their diff.
---
## Current changes
!`git diff HEAD`
## Instructions
Summarize the changes above in two or three bullet points, then list any risks you notice.Two parts to the file. The block at the top fenced by --- lines is the frontmatter. It is a tiny block of settings written in YAML. The only required line in there is description, which tells Claude what the skill does and when to load it.
Below the frontmatter is plain Markdown. That is the body Claude reads when the skill fires. The line !\git diff HEAD`` is a shell command that runs first, with its output dropped into the message Claude sees. Then Claude follows the instructions under "Instructions". That is it. No build step. No registration. Save the file and Claude can use it on the next prompt.
How Claude Decides to Run a Skill
Two paths get a skill into the conversation. You can call it by name, like /code-review, and the autocomplete list shows you every installed skill. Or you can let Claude pick. In a normal session, Claude only sees the descriptions of every installed skill, not the bodies. It reads the descriptions, decides which ones fit your prompt, and pulls in the full body of any that do. The body stays in context for the rest of the session. If context gets compacted later, the most recent invocation of each skill gets re-attached, capped at the first 5,000 tokens per skill and 25,000 tokens total. (Source: code.claude.com/docs/en/skills.)
This is why the description matters more than anything else in the file. Bad description, the skill never fires. Good description, Claude reaches for it the moment the topic lands.
There are two caps you should know about. The combined description plus when_to_use text gets truncated at 1,536 characters in the listing Claude sees. And the default budget for all skill listings together is 1% of the model's context window, configurable via a setting called skillListingBudgetFraction. Past that budget, lower-priority skills get cut. (Source: code.claude.com/docs/en/skills.)
Skills hot-reload. Edit the file mid-session and the change is live on the next turn.
12 Real Claude Skills People Run Today
Pulled from Anthropic's official set and the three highest-starred community repos on GitHub. Each row links to the upstream SKILL.md so you can read the source.
| # | Skill | Source | What it does | One-line use case |
|---|---|---|---|---|
| 1 | pdf | Anthropic (github.com/anthropics/skills) | Generate and process PDF documents | "Make a one-page PDF report from this CSV." |
| 2 | xlsx | Anthropic (github.com/anthropics/skills) | Create and edit Excel spreadsheets | "Turn this list into an Excel file with totals." |
| 3 | pptx | Anthropic (github.com/anthropics/skills) | Build PowerPoint presentations | "Generate a 10-slide deck from this brief." |
| 4 | mcp-builder | Anthropic (github.com/anthropics/skills) | Scaffold a new MCP server | "Build an MCP server that talks to my internal API." |
| 5 | skill-creator | Anthropic (github.com/anthropics/skills) | Help Claude write new skills | "I keep doing X by hand. Make a skill out of it." |
| 6 | webapp-testing | Anthropic (github.com/anthropics/skills) | Run automated browser tests | "Click through the signup flow and tell me what breaks." |
| 7 | tdd | mattpocock (github.com/mattpocock/skills) | Red-green-refactor TDD loop | "Add this feature using test-driven development." |
| 8 | caveman | mattpocock (github.com/mattpocock/skills) | Compress prose to roughly 75% fewer tokens | "Rewrite this in caveman so it costs less to send." |
| 9 | diagnose | mattpocock (github.com/mattpocock/skills) | Structured debugging methodology | "This bug is weird. Walk through it methodically." |
| 10 | code-review-and-quality | addyosmani (github.com/addyosmani/agent-skills) | Five-axis code review with severity labels | "Review this PR like a senior engineer would." |
| 11 | security-and-hardening | addyosmani (github.com/addyosmani/agent-skills) | OWASP Top 10 prevention checks | "Audit my app for the top web security risks." |
| 12 | systematic-debugging | obra (github.com/obra/superpowers) | Four-phase root-cause analysis | "Don't patch the symptom. Find the actual root cause." |
The three community repos behind rows 7 to 12 carry well over 80,000 stars each on GitHub, with obra/superpowers the most-starred of the bunch. (Sources: each repo's README on GitHub.)
Where Skills Live on Your Machine
Four locations, ranked by override order.
| Location | Path | Applies to |
|---|---|---|
| Enterprise | Managed settings | All users in the org |
| Personal | ~/.claude/skills/<name>/SKILL.md | Every project on your machine |
| Project | .claude/skills/<name>/SKILL.md | This project only |
| Plugin | <plugin>/skills/<name>/SKILL.md | Wherever the plugin is enabled |
Order is enterprise wins over personal, personal wins over project. Plugin skills carry a namespace like plugin-name:skill-name, so they cannot collide with anything you wrote yourself. (Source: code.claude.com/docs/en/skills.)
Skill vs Subagent vs MCP vs Plugin vs Project
This is the matrix the current search results are missing. Five things people confuse with each other, in one table.
| Need | Use this | Why |
|---|---|---|
| Teach Claude how to do something the same way every time | Skill | Static instructions Claude reads when the description matches. Cheap. Lives on disk. |
| Connect Claude to a live system like Stripe, Postgres, Slack | MCP server | MCP is the connection layer. It pulls live data and runs actions. |
| Spawn an isolated worker with its own context window and tools | Subagent | Subagents have their own context, system prompt, and tool list. |
| Bundle skills, MCP servers, commands, and hooks together to share | Plugin | A plugin is the distributable container around the other primitives. |
| Provide background reference for one workspace | Project | Projects hold knowledge files scoped to one workspace. |
Anthropic's own framing on the same question: "If you've got more skills than MCP servers, you're probably doing it right." (Source: claude.com/blog/skills-explained.)
The cleanest way to remember it. MCP brings live data in. Subagents fork off a worker. Plugins package the whole bundle for distribution. Projects hold reference material. Skills teach procedure.
What Else Goes Inside a Skill Folder
SKILL.md is the only required file. Most production skills also carry one or more of these:
scripts/for helper scripts the skill callsreferences/for documentation Claude can read on demandexamples/for sample inputs and outputstemplates/for files the skill copies and fills in
None of those are loaded into context by default. Claude opens them only when the body of SKILL.md tells it to. That is how a skill can carry hundreds of pages of reference without burning tokens until it needs them.
The Frontmatter Fields, in Plain English
Every field except description is optional. Here are the ones you will actually use, with what they do:
| Field | Required | What it does |
|---|---|---|
name | No | Display name. Defaults to the directory name. Lowercase, numbers, hyphens. Max 64 characters. |
description | Recommended | What the skill does and when to use it. Claude reads this to decide whether to load the skill. |
when_to_use | No | Extra trigger phrases. Appended to the description in the listing. |
argument-hint | No | Autocomplete hint, like [issue-number]. |
arguments | No | Named positional arguments for $name substitution. |
disable-model-invocation | No | If true, only the user can call the skill. Claude will not auto-fire it. |
user-invocable | No | If false, hides the skill from the / menu. For background skills only. |
allowed-tools | No | Tools the skill can use without a per-call approval. |
model | No | Override the model for the rest of the turn. |
effort | No | Effort level: low, medium, high, xhigh, max. |
context | No | Set to fork to run the skill in a fresh subagent context. |
agent | No | Which subagent type to use when context: fork. |
hooks | No | Skill-scoped hooks. |
paths | No | Glob patterns. The skill auto-loads only when files matching these are open. |
shell | No | bash (default) or powershell. |
(Source: code.claude.com/docs/en/skills and resources.anthropic.com/hubfs/The-Complete-Guide-to-Building-Skill-for-Claude.pdf.)
If you only learn one field, learn description. It is the entire reason Claude does or does not pick your skill.
What Changed in 2026
The skills format moved fast through the spring. Three things worth knowing as of May 2026.
Root SKILL.md auto-surface. Around Claude Code v2.1.69 in late April 2026, plugins with a root-level SKILL.md and no skills/ subdirectory now show up as a single skill. Single-file plugin distribution is no longer awkward.
Custom commands merged into skills. A file at .claude/commands/deploy.md and a skill at .claude/skills/deploy/SKILL.md both create /deploy and behave the same way. Old .claude/commands/ files keep working. The distinction between the two has dissolved.
An open standard for skills. The format is published at agentskills.io, so the same SKILL.md file works across more than just Claude. Community marketplaces like claudemarketplaces.com and tonsofskills.com collectively list more than 4,000 skills as of May 2026. (These are community-run, not Anthropic-run; the official source remains github.com/anthropics/skills.)
Bundled skills now ship with Claude Code itself: /simplify, /batch, /debug, /loop, /claude-api, /init, /review, /security-review. You already have those even if you have never installed a thing.
When a Skill Refuses to Fire
The most common reason: Claude never picked your skill, because the description did not match the prompt the user wrote. Three fixes.
Rewrite the description to lead with what the skill does, then add explicit example phrases the user might say. The 1,536-character cap is generous, so use it.
Run /doctor to confirm Claude is seeing the skill at all. The output lists every skill loaded in the current session, with their descriptions.
If you have hundreds of skills installed and most are getting cut from the listing, raise skillListingBudgetFraction past the default 1%. (Source: code.claude.com/docs/en/skills.)
Triggering is probabilistic, not deterministic. Claude usually picks the right skill. Sometimes it does not. The description is the lever.
Want 95 Skills Without Writing One
Build This Now ships with 95+ Claude Skills already wired up. Planning, building, testing, security, design, deployment, the lot. Drop the framework into a Next.js plus Supabase project and Claude inherits every one of them on day one. From idea to SaaS in 48 hours, $79 one-time, no subscription.
That is the gap most people hit between "I wrote my first ten-line skill" and "I have a full production build system on tap."
Where to Go Next
Three follow-ups worth your time. Read the official spec at code.claude.com/docs/en/skills. Browse the 17 official skills at github.com/anthropics/skills. Then run the skill-creator skill the next time you catch yourself doing the same task by hand twice.
A skill is a folder. A folder is a file. A file is a description plus a recipe. Build the folder once. Claude uses it forever.
Posted by @speedy_devv
Arrêtez de configurer. Commencez à construire.
Templates SaaS avec orchestration IA.
Le répertoire de règles Claude Code
Le répertoire .claude/rules découpe CLAUDE.md en fichiers markdown ciblés par chemin. Chaque fichier se charge uniquement là où il s'applique, avec la même priorité élevée que CLAUDE.md.
Guide des Skills Claude
Les skills sont des dossiers d'instructions que Claude Code charge à la demande. Dépose SKILL.md dans .claude/skills/ton-skill pour les workflows procéduraux, les checklists, les règles maison.