Skip to content

Quick Start

This guide will help you get started with AI programming core concepts in 5 minutes.

Prerequisites

  • Basic programming knowledge (any language)
  • Basic understanding of AI programming assistants (e.g., Claude Code, Cursor)
  • Node.js 18+ installed (if learning OpenSpec)

1. Understand Core Concepts

Before starting, let's quickly understand the three core concepts:

What is a Skill?

A Skill is a reusable AI instruction template. It tells the AI assistant how to complete a specific task, including:

  • name: Skill name
  • description: Function description
  • prompt: Detailed execution instructions
  • examples: Input/output examples

One-sentence summary: Skill = Standardized work manual for AI

What is OpenSpec?

OpenSpec is a Spec-Driven Development (SDD) framework. Core idea:

Define specifications before writing code. Ensure humans and AI agree on requirements.

Workflow:

  1. /opsx:propose - Create proposal
  2. /opsx:apply - Apply specifications, AI implements code
  3. /opsx:archive - Archive completion

What is Harness?

Harness is a testing infrastructure for evaluating AI output quality:

  • Test Scenarios: Define inputs and expected outputs
  • Evaluators: Automated evaluation of AI outputs
  • Mock Server: Test environment without real API calls

2. Create Your First Skill

Create a simple code explanation Skill:

json
{
  "name": "explain-code",
  "description": "Explain how code works with detailed analysis",
  "prompt": "Please explain how the following code works:\n\n1. Overall functionality\n2. Key variables and functions\n3. Execution flow\n4. Potential optimization suggestions\n\nCode:\n{{code}}",
  "examples": [
    {
      "input": "def fib(n):\n    if n <= 1:\n        return n\n    return fib(n-1) + fib(n-2)",
      "output": "This is a recursive Fibonacci function implementation..."
    }
  ]
}

Save the file as explain-code.json in your project's .skills/ directory.

3. Experience OpenSpec Workflow

Install OpenSpec CLI:

bash
npm install -g @fission-ai/openspec@latest

Initialize project:

bash
cd your-project
openspec init

Create your first change:

bash
/opsx:propose "Add user login feature"

AI will generate:

  • proposal.md - Proposal description
  • design.md - Technical design
  • specs/ - Detailed requirements
  • tasks.md - Implementation tasks

After review and confirmation, let AI implement:

bash
/opsx:apply

4. Next Steps

Congratulations! You've learned the three core paradigms of AI programming. Next:

FAQ

Q: Which one should I learn first?

A: Suggested order: Skill → OpenSpec → Harness. Skill is the foundation, OpenSpec is the workflow, Harness is quality assurance.

Q: Do these tools depend on a specific AI?

A: No. Skills are a universal format, OpenSpec and Harness are decoupled from specific AI tools. You can use them in any tool that supports these formats.

Q: Is the learning curve steep?

A: Getting started is easy! You can create your first Skill in 5 minutes. Going deep requires practice, but there's a clear learning path at every step.

基于 MIT 许可发布