- Claude's "skills" are Markdown files that automate repetitive instructions, teaching the AI how to perform tasks or apply preferences once.
- Skills are automatically invoked when Claude's AI recognizes a match between your request and a skill's description, eliminating manual repetition.
- They are ideal for standardizing specialized knowledge like coding styles, commit formats, or brand guidelines across personal or team projects.
What are skills?
- Define Skills with Markdown: Create skills as Markdown files to instruct Claude on specific procedures, formats, or standards.
- Automatic Activation: Claude automatically activates relevant skills by matching your request against their descriptions, without requiring explicit commands.
- Personal Skills Location: Store your individual preferences and styles (e.g., commit message format) in
home/.Claude/skillsto follow you across projects. - Project Skills for Teams: Place team-specific standards (e.g., company brand guidelines) in
.Claude/skillswithin a repository's root, making them available to all collaborators. - Context Efficiency: Skills load only their name and description on demand, preserving the context window, unlike
CLAUDE.mdfiles which are always present. - When to Use Skills: Implement skills for specialized, frequently repeated instructions that pertain to specific tasks, such as code review checklists or documentation formats.
Claude— An AI assistant that can be taught specific behaviors and standards.Skill— A Markdown file that teaches Claude how to perform a specific task or apply a preference automatically.Agent skills— Folders containing instructions, scripts, and resources for AI agents to discover and use.PR review— The process of evaluating code changes submitted in a Pull Request (PR) before they are merged.Commit message— A brief text description accompanying a set of code changes that are saved (committed) to a version control system.Context window— The limited amount of previous conversation or information an AI model can reference at any given time.Slash commands— Explicit commands initiated by a/symbol that a user types to trigger a specific AI action.Repository— A digital storage location for all the files, folders, and revision history of a project, typically managed by a version control system like Git.
Every time you explain your team's coding standards to Claude, you're repeating yourself. Every PR review, you redescribe how you want feedback [music] structured. Every commit message, you remind Claude of your preferred format, and skills fix this. A skill is a markdown file that teaches Claude [music] how to do something once, and Claude applies that knowledge automatically whenever it's relevant. Agent skills are folders of instructions, scripts, and resources that Asia can discover and use to do things more accurately and efficiently. With Claw Code, we have the skill MD file. The description is how Claw decides whether to use the skill. When you ask Claude to review this PR, it matches your request against available skill descriptions and finds this one. Claude reads your request, compares it to all available skill descriptions, and activates the ones that match. You can store skills in a few places depending on who needs them. Personal skills go in the home directory.Claude/skills and follow you across all your project. These are your preferences, your commit message style, your documentation format, how you like code explained. Project skills go in the do.claw/skills inside of the root directory of your repository. Anyone who clones the repository gets these skills automatically. This is where team standards live, like your company's brand guidelines, preferred fonts, and colors that you use for web design. Claude Code has several ways to customize behavior. Skills are unique because they're automatic and task specific. CloudMD files load into every conversation. If you want claw to always use typescript strict mode that goes in your CLAUDE.md file. Skills on the other hand load on demand when they match your request. It only loads in the name and description. So it doesn't fill up your entire context window. Your PR review checklist doesn't need to be in the context when you're debugging. It loads when you actually ask for a review. Slash commands require you to type them. Skills don't. Claude applies them when it recognizes the situation. Skills work best for specialized knowledge that applies to specific tasks, code review standards your team follows, commit message formats that you prefer, brand guidelines of your organization. If you find yourself explaining the same thing [music] to Claude repeatedly, well, that's a skill waiting to be written.
TL;DR
- Claude's "skills" are Markdown files that automate repetitive instructions, teaching the AI how to perform tasks or apply preferences once.
- Skills are automatically invoked when Claude's AI recognizes a match between your request and a skill's description, eliminating manual repetition.
- They are ideal for standardizing specialized knowledge like coding styles, commit formats, or brand guidelines across personal or team projects.
Takeaways
- Define Skills with Markdown: Create skills as Markdown files to instruct Claude on specific procedures, formats, or standards.
- Automatic Activation: Claude automatically activates relevant skills by matching your request against their descriptions, without requiring explicit commands.
- Personal Skills Location: Store your individual preferences and styles (e.g., commit message format) in
home/.Claude/skillsto follow you across projects. - Project Skills for Teams: Place team-specific standards (e.g., company brand guidelines) in
.Claude/skillswithin a repository's root, making them available to all collaborators. - Context Efficiency: Skills load only their name and description on demand, preserving the context window, unlike
CLAUDE.mdfiles which are always present. - When to Use Skills: Implement skills for specialized, frequently repeated instructions that pertain to specific tasks, such as code review checklists or documentation formats.
Vocabulary
Claude— An AI assistant that can be taught specific behaviors and standards.Skill— A Markdown file that teaches Claude how to perform a specific task or apply a preference automatically.Agent skills— Folders containing instructions, scripts, and resources for AI agents to discover and use.PR review— The process of evaluating code changes submitted in a Pull Request (PR) before they are merged.Commit message— A brief text description accompanying a set of code changes that are saved (committed) to a version control system.Context window— The limited amount of previous conversation or information an AI model can reference at any given time.Slash commands— Explicit commands initiated by a/symbol that a user types to trigger a specific AI action.Repository— A digital storage location for all the files, folders, and revision history of a project, typically managed by a version control system like Git.
Transcript
Every time you explain your team's coding standards to Claude, you're repeating yourself. Every PR review, you redescribe how you want feedback [music] structured. Every commit message, you remind Claude of your preferred format, and skills fix this. A skill is a markdown file that teaches Claude [music] how to do something once, and Claude applies that knowledge automatically whenever it's relevant. Agent skills are folders of instructions, scripts, and resources that Asia can discover and use to do things more accurately and efficiently. With Claw Code, we have the skill MD file. The description is how Claw decides whether to use the skill. When you ask Claude to review this PR, it matches your request against available skill descriptions and finds this one. Claude reads your request, compares it to all available skill descriptions, and activates the ones that match. You can store skills in a few places depending on who needs them. Personal skills go in the home directory.Claude/skills and follow you across all your project. These are your preferences, your commit message style, your documentation format, how you like code explained. Project skills go in the do.claw/skills inside of the root directory of your repository. Anyone who clones the repository gets these skills automatically. This is where team standards live, like your company's brand guidelines, preferred fonts, and colors that you use for web design. Claude Code has several ways to customize behavior. Skills are unique because they're automatic and task specific. CloudMD files load into every conversation. If you want claw to always use typescript strict mode that goes in your CLAUDE.md file. Skills on the other hand load on demand when they match your request. It only loads in the name and description. So it doesn't fill up your entire context window. Your PR review checklist doesn't need to be in the context when you're debugging. It loads when you actually ask for a review. Slash commands require you to type them. Skills don't. Claude applies them when it recognizes the situation. Skills work best for specialized knowledge that applies to specific tasks, code review standards your team follows, commit message formats that you prefer, brand guidelines of your organization. If you find yourself explaining the same thing [music] to Claude repeatedly, well, that's a skill waiting to be written.