What is the Human-AI Document Standard (HADS)?

HADS is a structured Markdown convention designed for documentation that is read by both humans and AI models, using explicit tags like [SPEC] and [NOTE] alongside an AI reading manifest.

Can I convert existing README or API documentation to HADS?

Yes, this Claude Code skill can automatically parse existing markdown docs, extracting facts into [SPEC] blocks, moving context to [NOTE] blocks, and generating standard headers and manifests.

What are the mandatory elements of a valid HADS document?

A valid HADS document requires an H1 title, a version declaration in the top 20 lines, an AI Reading Instruction manifest before the first section, and properly formatted bold tags.

How does HADS optimize token usage for AI coding assistants?

HADS includes an AI reading manifest that instructs models to focus primarily on authoritative [SPEC] and [BUG] blocks while skipping non-essential narrative [NOTE] blocks unless extra context is needed.

HADS Documentation Standard - Claude Code Skill Guide

The HADS skill standardizes technical documentation for dual human-AI consumption using structured block types like [SPEC], [NOTE], [BUG], and [?]. It enables Claude to generate, validate, and convert standard markdown into token-optimized HADS format with explicit AI reading instructions. This skill is ideal for technical writers, software engineers, and documentation maintainers looking to make codebase documentation instantly interpretable and token-efficient for LLM agents while preserving rich narrative context for human readers.

주요 기능

01Automated AI Manifest generation directing models on what to read and skip

02Token-efficient scanning and parsing for large technical documentation files

03Structured block tagging ([SPEC], [NOTE], [BUG], [?]) to separate core facts from narrative context

04Syntax and structure validation for HADS compliance and required metadata

05Conversion of standard Markdown, READMEs, and API docs into HADS format

0639,277 GitHub stars

사용 사례

01Drafting new software specifications with structured failure modes and authoritative facts

02Converting existing project READMEs and API references into dual-purpose human-AI documentation

03Optimizing documentation context windows and token usage for AI coding assistants and agents

Tieline — product intent grounded in code

HADS Documentation Standard

주요 기능

사용 사례

Diagram Maker & Visualizer

Iso 14001 Internal Audit

Next.js Turbopack Optimizer

주요 기능

사용 사례