LLM Context Files

import { Aside } from ‘@astrojs/starlight/components’;

Cyntex provides standard llms.txt files so any LLM can quickly understand Cyntex.

Quick Start

Provide the following URLs to your AI model so it can understand Cyntex and help you write .cyn.yml:

File	URL	Purpose
llms.txt	https://docs.cyntex.io/llms.txt	Condensed version (concepts + quick reference); suitable for most scenarios
llms-full.txt	https://docs.cyntex.io/llms-full.txt	Full version (all DSL syntax details); suitable for complex tasks

Using with Claude

Please read https://docs.cyntex.io/llms.txt and then help me create a CDC sync task from MySQL to MongoDB

Using with ChatGPT / GPT-4o

Use the “Browse” feature to fetch https://docs.cyntex.io/llms.txt, then ask your question.

Using with Any AI That Supports URLs

Paste the URL directly into the conversation. Most modern LLMs can fetch and understand file content.

---

JSON Schema

Cyntex’s DSL schema is particularly AI-friendly — every field includes description and examples, generated from the same source as the code:

https://docs.cyntex.io/schema/v1/pipeline.json
https://docs.cyntex.io/schema/v1/source.json

IDE Integration

Add the following to your VS Code settings.json:

{
  "yaml.schemas": {
    "https://docs.cyntex.io/schema/v1/pipeline.json": "*.cyn.yml"
  }
}

---

Example Corpus

https://docs.cyntex.io/examples/

Contains valid .cyn.yml samples covering all scenarios defined in ADR-0016 §14, ready to be provided directly to an AI as few-shot examples.

---

MCP Integration

The MCP server is enabled in the **Alpha phase**. The POC phase has only the offline CLI.

When the Cyntex service is running, an AI agent can operate directly via the MCP protocol:

mcp://localhost:7778

This lets the AI not only write .cyn.yml but also deploy, start, and monitor tasks. See MCP Integration.