Claude Skill Generator
Describe a task you keep repeating in plain English and get back a complete SKILL.md you can drop straight into Claude Code, with the name and description checked live against the real Agent Skills constraints as you type. The generator writes the description in a what-it-does-and-when-to-use-it shape, since that field is the only thing Claude reads when deciding whether your skill applies to a request. It also flags a draft live when it looks like it covers more than one unrelated job, since a narrowly scoped skill triggers far more reliably than a broad one.
1. Describe the task you keep repeating
Plain language is fine. The more specific the workflow, the sharper the generated skill.
Or start from one of these
Your SKILL.md will appear here, fully editable, with the name and description checked against Anthropic's real constraints as you type.
What actually goes in a SKILL.md
A skill is just a folder with a SKILL.md inside it: YAML frontmatter on top, markdown instructions below. Only two frontmatter fields are required, and both have hard rules this tool enforces while you type. name is capped at 64 characters, allows lowercase letters, numbers and hyphens only, and cannot contain the reserved words “claude” or “anthropic”. description must be non-empty and under 1024 characters - and it is the single field worth agonising over, because it is the text Claude reads to decide whether your skill is relevant to a request at all.
That is why every good description names two things: what the skill does and when it should fire. “Reviews a pull request diff” on its own gives Claude nothing to match against. “Reviews a pull request diff and flags risky changes. Use when the user asks to review a PR or check a diff before merging” tells it exactly which requests belong to this skill. The generator writes descriptions in that shape, and flags yours if the trigger half is missing.
Below the frontmatter it is ordinary markdown - a short overview, an ## Instructions list of steps, and examples where they earn their place. Bigger skills can also reference extra bundled files next to SKILL.md (a REFERENCE.md, a scripts/ folder Claude runs via bash) so the heavy detail is only pulled in when it is actually needed.
Frequently asked questions
What is a Claude skill, exactly?
A skill is procedural knowledge you hand Claude once and reuse forever: a directory containing a SKILL.md file, made of YAML frontmatter (a name and a description) plus a markdown body of instructions. When a request matches the description, Claude loads those instructions and follows them. It is the difference between re-typing the same three-paragraph prompt every week and just typing a slash command. A skill can also reference extra files beside it - a FORMS.md, a REFERENCE.md, or scripts under a scripts/ folder that Claude runs via bash, where only the output enters the conversation, not the script's code.
How do I write a Claude skill description that actually triggers?
State what the skill does and when Claude should use it, in that order, in plain third-person language. The description is the only thing Claude matches a user's request against when deciding whether to run your skill, so a vague one means the skill sits there unused. Name the concrete triggers you expect - 'Use when the user asks to review a PR, check a diff, or asks what looks risky in these changes' - and stay under the 1024-character limit. This SKILL.md generator drafts the description in that what-plus-when shape and warns you live if an edit drops the trigger half.
Is a skill the same thing as a custom slash command?
They overlap. Claude Code has unified the two: a command file and a skill both give you the same /<name> behaviour and accept the same frontmatter, and both support the $ARGUMENTS placeholder system. The difference is structure - a skill is a folder, so it can carry supporting files, scripts and progressive-disclosure references, and it takes a name field of its own (a command file ignores name and paths, because its file name is the command name). A command is a single markdown file, faster to write when the job is one short instruction. Reach for this Claude Code custom skill generator when the workflow has real steps, and the slash command generator when it does not.
Should I really split one big skill into several smaller ones?
In most cases, yes - this is the single most common real-world mistake in Agent Skills. A skill trying to cover several unrelated responsibilities (say, handling commits, writing PR descriptions, and updating a changelog all in one) triggers unreliably, because its description has to vaguely cover all three jobs at once, and Claude cannot tell which part applies to a given request. One well-known review of 47 popular community skills found forty of them actually made Claude's output worse than not using a skill at all, largely for this reason. This generator flags the draft live when a description sounds like it is covering more than one job, and recommends the split.
Building something short and one-line instead? The AI Slash Command Generator writes the single-file version - same /command, less scaffolding.