LLMs

Generates llms.txt and llms-full.txt indexes following the llmstxt.org spec—a standard for making websites and codebases navigable by LLMs.

What This Skill Does

LLMs creates machine-readable indexes of your documentation or codebase. llms.txt is a concise table of contents with links. llms-full.txt expands each link with full inline content so LLMs can consume the entire site in one context load.

When to Use

Publishing a docs site that should be AI-accessible
Creating a codebase overview for LLM context injection
Building an API or SDK that developers use with AI tools
Improving AI assistant accuracy when working with your project

Core Capabilities

Crawl local path or remote URL
Generate llms.txt (compact index) and llms-full.txt (expanded inline content)
Follow llmstxt.org spec exactly
Custom output path

Arguments

Argument	Description
`[path\|url]`	Local directory or remote URL to index (default: current dir)
`--full`	Also generate `llms-full.txt` with expanded inline content
`--output path`	Output directory (default: project root)

Example Usage

/ck:llms
/ck:llms ./docs --full
/ck:llms https://docs.mysite.com --full --output ./public
/ck:llms src/content/docs --output ./public

Output Format

llms.txt follows the llmstxt.org spec:

# Project Name

> One-line description of the project

## Docs

- [Quick Start](/docs/getting-started/quick-start): Get up and running in 5 minutes
- [API Reference](/docs/api): Complete API documentation

## Optional

- [Changelog](/docs/changelog): Version history

llms-full.txt includes the same structure but with full page content inlined under each link.

Docs — manages documentation this skill indexes
Repomix — alternative for packing repos into AI-friendly context
Docs Seeker — retrieves docs for consumption