hubify generate

Generate a new skill from a natural language prompt using AI.

Usage

hubify generate <prompt> [options]

Arguments

Argument	Description
`prompt`	Natural language description of the skill

Options

Option	Description
`--output <path>`	Output directory (default: current)
`--name <name>`	Skill name (auto-generated if not provided)
`--category <cat>`	Skill category
`--interactive`	Interactive refinement mode
`--model <name>`	AI model to use

Examples

Basic Generation

hubify generate "Python code review checklist with security focus"

Output:

Generating skill from prompt...

Analyzing requirements...
  - Category: security
  - Platforms: all
  - Use cases: code review, security audit

Generating skill content...

Created: python-security-review/.hub

  Name: python-security-review
  Version: 0.1.0
  Category: security

  Use Cases:
    - Review Python code for security vulnerabilities
    - Check for injection attacks
    - Validate input handling
    - Audit dependency security

Edit the .hub file to refine, then publish with:
  hubify publish ./python-security-review

Specify Output Path

hubify generate "React component testing patterns" --output ./my-skills

Specify Name

hubify generate "TypeScript API client patterns" --name ts-api-client

Interactive Mode

hubify generate "Database migration best practices" --interactive

Interactive mode allows you to:

Refine the generated prompt
Add/remove use cases
Adjust examples
Review before saving

Specify Category

hubify generate "CI/CD pipeline setup" --category devops

Generated Structure

python-security-review/
├── .hub              # Skill manifest
├── prompt.md         # Generated prompt
└── examples/
    └── basic.md      # Example usage

.hub File

The generated .hub file:

name: python-security-review
version: 0.1.0
category: security
author: "@yourusername"

description: |
  Security-focused Python code review checklist

platforms:
  - claude-code
  - cursor
  - windsurf

use_cases:
  - Review Python code for security vulnerabilities
  - Check for SQL injection
  - Validate user input handling
  - Audit third-party dependencies

tags:
  - python
  - security
  - code-review
  - owasp

prompt_file: ./prompt.md

examples:
  - title: "Flask API Review"
    prompt: "Review this Flask API for security issues"
    expected: "Comprehensive security audit with findings"

Refining Generated Skills

Edit the Prompt

# Open in editor
code python-security-review/prompt.md

Validate Changes

hubify hub validate ./python-security-review

Test Locally

hubify test python-security-review

Publish

hubify publish ./python-security-review

Tips

Be Specific

# Good
hubify generate "React hooks patterns for data fetching with error handling and loading states"

# Less specific
hubify generate "React patterns"

Include Context

hubify generate "TypeScript configuration for Next.js 14 with app router and strict mode"

Mention Platforms

hubify generate "Git commit message format optimized for Claude Code"

Model Selection

Choose the AI model for generation:

hubify generate "..." --model gpt-4
hubify generate "..." --model claude-3

Batch Generation

Generate multiple related skills:

hubify generate "React testing patterns" --name react-unit-tests
hubify generate "React integration testing" --name react-integration-tests
hubify generate "React E2E testing with Playwright" --name react-e2e-tests