Skip to main content

AI Helper Overview

This guide helps you leverage AI coding agents (Claude Code, Cursor, GitHub Copilot, Windsurf, etc.) to accelerate your CVT integration.

What AI Agents Can Help With​

AI coding agents excel at CVT-related tasks:

  • Setup & Configuration: Installing SDKs, configuring the server, setting up Docker
  • Writing Contract Tests: Creating validation tests for your API consumers
  • Debugging Validation Errors: Understanding why requests/responses fail validation
  • Schema Comparison: Interpreting breaking change reports
  • CI/CD Integration: Setting up contract testing in your pipelines
  • Fixture Generation: Creating schema-compliant test data

The llms.txt Approach​

CVT provides LLM-friendly documentation through standardized llms.txt files. These files are optimized for AI agents to quickly understand the project and provide accurate assistance.

Available Files​

FilePurposeBest For
llms.txtSummary with linksQuick orientation, finding specific docs
llms-full.txtComplete API referenceDetailed implementation help

How to Use​

Point your AI agent to the llms.txt files when starting a CVT-related task:

I'm integrating with CVT for contract testing.
Reference: https://raw.githubusercontent.com/sahina/cvt/main/llms.txt

For detailed implementation questions:

I need help with CVT's consumer registration API.
Reference: https://raw.githubusercontent.com/sahina/cvt/main/llms-full.txt

Supported AI Tools​

These tools work well with CVT's documentation:

Claude Code​

Add the llms.txt reference to your CLAUDE.md file. See Context Templates for the full template.

Cursor​

Add the llms.txt reference to your .cursorrules file. See Context Templates for the full template.

GitHub Copilot​

Reference llms.txt in your copilot instructions file or conversation context.

Windsurf​

Add llms.txt to your project context configuration.

Other Tools​

Any AI coding agent that supports project context files can use the llms.txt approach.

If your AI tool supports agent skills (Claude Code, Codex CLI, Gemini CLI, Cursor, and others), CVT provides 6 agent skills in .agents/skills/ that guide workflows step-by-step with auto-language-detection and SDK-specific code:

SkillWhat It Does
/cvt-setupInstall SDK, configure server, create first contract test
/cvt-write-contractWrite consumer contract tests
/cvt-producer-testValidate API responses against your schema
/cvt-ciIntegrate CVT into CI/CD pipelines
/cvt-breaking-changesDetect breaking schema changes
/cvt-troubleshootDiagnose common CVT issues

Install​

Run this from your project root to copy the skills into your project:

curl -sL https://github.com/sahina/cvt/archive/main.tar.gz \
  | tar xz --strip-components=1 'cvt-main/.agents'

Commit .agents/skills/ to version control so your whole team gets the skills. To update, re-run the same command.

Usage​

  1. Open your project in an agent-supported editor (Claude Code, Codex CLI, Gemini CLI, Cursor, etc.)
  2. The agent discovers skills automatically from .agents/skills/
  3. Ask the agent to run /cvt-setup to get started

All skills support Node.js, Python, Go, and Java SDKs with automatic language detection.

llms.txt Approach (Universal)​

For agents that don't support SKILL.md, or for quick reference, CVT provides LLM-friendly documentation:

  1. Add context to your project: Create a context file for your AI tool (see Context Templates)

  2. Point to llms.txt: Include the URL in your context file:

    https://raw.githubusercontent.com/sahina/cvt/main/llms.txt

  3. Start asking questions: Your AI agent now has CVT knowledge

Example Conversation​

User: Set up CVT for my Node.js project

AI: I'll help you set up CVT. Based on the CVT documentation, here's what we need:

1. Start the CVT server:
   docker run -d -p 9550:9550 ghcr.io/sahina/cvt:latest

2. Install the Node.js SDK (see installation guide for details):
   npm install @sahina/cvt-sdk

3. Create your first contract test...

What's in This Guide​

PageContent
Context TemplatesCopy-paste templates for CLAUDE.md, .cursorrules, and other tools
Common MistakesPitfalls to avoid when using AI with CVT
Advanced PatternsSchema evolution, CI/CD, and fixture generation workflows
OpenAPI Schema GeneratorAdaptive prompt for generating or improving OpenAPI schemas from code

Next Steps​