How does this help AI agents understand my code better?

By enforcing 'self-contained' writing rules and adding AI-specific metadata, agents can understand documentation chunks in isolation without needing to read the entire site to gain context.

What is the purpose of the llms.txt file?

The llms.txt file acts as a navigation index for AI agents, providing a compact map of all available documentation and its purpose in a format LLMs can easily process.

What is the Dual-Audience documentation model?

It is a strategy where a single Markdown source file produces two distinct outputs: visual HTML for human readers and structured, context-rich Markdown for AI agents.

Do I have to write my documentation twice?

No. You write your documentation once in Markdown/MDX, and the skill ensures the formatting and metadata satisfy both human and AI requirements simultaneously.

Does this skill require a specific tech stack?

Yes, it is designed for Astro websites using the Starlight documentation framework and the starlight-llms-txt plugin.

Dual-Audience Documentation Manager

Name: Dual-Audience Documentation Manager
Author: kormie

bykormie

0•

内容管理

Authorizes documentation that simultaneously serves human readers through rendered HTML and AI agents via optimized Markdown.

This skill enables developers to implement the dual-audience documentation model within Astro and Starlight projects, ensuring content is perfectly tailored for both people and machines. By leveraging the starlight-llms-txt plugin, it automates the creation of machine-readable files like llms.txt while enforcing strict writing conventions—such as self-contained sections, explicit noun usage, and AI-specific frontmatter—that prevent context loss when AI agents retrieve documentation chunks. It is an essential tool for projects that want to be 'AI-ready' without maintaining separate documentation tracks.

主要功能

01Standardizes AI-specific frontmatter including content-type, scope, and dense ai-summaries.

02Generates machine-optimized llms.txt, llms-full.txt, and llms-small.txt files automatically.

03Enforces self-contained section writing to ensure AI agents can parse documentation in isolated chunks.

04Synchronizes project terminology using a centralized configuration to prevent linguistic ambiguity.

050 GitHub stars

06Optimizes heading hierarchies and code block tagging for better semantic mapping by LLMs.

使用场景

01Transforming existing Starlight documentation sites into AI-optimized knowledge bases using the llms.txt standard.

02Building documentation for libraries that need to be easily indexed and understood by AI coding assistants.

03Maintaining internal technical manuals that require both high visual quality for teams and high retrievability for RAG systems.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add kormie/dot-skills dual-docs

For use in Claude.ai and ChatGPT

Download Skill