CLI for creating Claude Agent Skills with progressive disclosure validation. Built for Claude Code to use when humans ask it to create skills. Enables operations teams to automate skill creation with specific package manager commands.
git clone https://github.com/spences10/claude-skills-cli.gitclaude-skills-cli is a command-line tool designed for Claude Code to programmatically generate Claude Agent Skills with built-in progressive disclosure validation. It enforces a three-level skill architecture that optimizes token usage by loading metadata, documentation, and references on-demand rather than all at once. The CLI provides initialization, validation, and packaging commands with configurable strictness modes, helping teams standardize skill structure and prevent oversized documentation. It includes activation hooks that improve skill reliability in Claude Code, addressing the common issue of skills failing to auto-activate without explicit instructions.
[{"step":"Install the claude-skills-cli package globally using your preferred package manager. For npm users: 'npm install -g claude-skills-cli'. For pip users: 'pip install claude-skills-cli'.","tip":"Ensure you have Node.js (v18+) or Python (v3.10+) installed before installation."},{"step":"Initialize a new skill project by running 'claude-skills-cli init [SKILL_NAME]'. Replace [SKILL_NAME] with your skill's identifier (e.g., 'customer-onboarding').","tip":"Use kebab-case for skill names to ensure compatibility with all systems. Example: 'data-pipeline-monitor' instead of 'DataPipelineMonitor'."},{"step":"Edit the generated skill.json file to define your input fields, validation rules, and progressive disclosure steps. Use the provided templates in the ./skills/[SKILL_NAME] directory.","tip":"Start with the prompt_template.md file to define the user-facing prompt structure. Keep progressive steps simple (3-5 steps max) for better user experience."},{"step":"Run validation tests locally using 'claude-skills-cli test ./skills/[SKILL_NAME]'. Fix any validation errors before proceeding.","tip":"Use the --verbose flag to get detailed error messages: 'claude-skills-cli test ./skills/[SKILL_NAME] --verbose'."},{"step":"Publish your skill to your Claude Code environment using 'claude-skills-cli publish ./skills/[SKILL_NAME]'. The CLI will handle package bundling and dependency resolution.","tip":"For CI/CD pipelines, use the --no-install flag to skip dependency installation: 'claude-skills-cli publish ./skills/[SKILL_NAME] --no-install'."}]
Automating skill creation for API client integrations via Claude Code prompts
Validating skill documentation structure before deployment to Claude.ai
Ensuring consistent progressive disclosure across multiple team-created skills
Packaging and uploading validated skills to Claude.ai
No install command available. Check the GitHub repository for manual installation instructions.
git clone https://github.com/spences10/claude-skills-cliCopy the install command above and run it in your terminal.
Launch Claude Code, Cursor, or your preferred AI coding agent.
Use the prompt template or examples below to test the skill.
Adapt the skill to your specific use case and workflow.
Use the claude-skills-cli to create a new skill for [SKILL_NAME] that validates user inputs progressively. The skill should include [INPUT_FIELDS] and enforce [VALIDATION_RULES]. Generate the skill files in the ./skills/[SKILL_NAME] directory. After creation, run the validation tests using 'claude-skills-cli test ./skills/[SKILL_NAME]'.
``` ✅ Skill 'inventory-audit' created successfully in ./skills/inventory-audit 📁 Generated files: - ./skills/inventory-audit/skill.json - ./skills/inventory-audit/prompt_template.md - ./skills/inventory-audit/validation_tests.json - ./skills/inventory-audit/README.md 🔍 Validation results: ✅ All 12 progressive disclosure tests passed ✅ Input field 'product_id' validated against SKU format ✅ Input field 'quantity' validated against warehouse limits ✅ Input field 'location' cross-validated with warehouse zones ✅ CLI command 'claude-skills-cli publish ./skills/inventory-audit' ready for deployment 📊 Skill metrics: - Input fields: 3 (product_id, quantity, location) - Validation rules: 5 (format, range, cross-reference, required, conditional) - Progressive steps: 4 (init → validate → confirm → submit) - Estimated skill size: 1.2MB ```
AI assistant built for thoughtful, nuanced conversation
IronCalc is a spreadsheet engine and ecosystem
ITIL-aligned IT service management platform
Customer feedback management made simple
Enterprise workflow automation and service management platform
Automate your spreadsheet tasks with AI power
Take a free 3-minute scan and get personalized AI skill recommendations.
Take free scan