Skip to main content
AI Inspiration World
AI AppsAI SkillsAI GamesAI Short Films

⌕
EN|中文
中
AI AppsAI SkillsAI GamesAI Short Films
Home/AI Skills/doc-coauthoring
AI Skills · Writingdoc-coauthoring

Added Oct 14, 2026

doc-coauthoring

An Anthropic example Skill that walks you through writing proposals, specs and decision docs with Claude in three stages, ending with a test read by a fresh Claude.

Ecosystem
Agent Skills
Skill category
Writing
Creator / organization
Anthropic
View work and source ↗

What it is

doc-coauthoring is an example Skill that Anthropic publishes in its official anthropics/skills repository. It guides you through writing documentation, proposals, technical specs, decision docs and similar structured content with Claude, and also activates when you mention a PRD, design doc or RFC. At the start, Claude explains the three stages and asks whether you want to follow the workflow or write freeform. Stage one gathers context: Claude asks about the document type, the main audience, the impact you want it to have, any template, and other constraints, invites you to dump all the background you have, then asks 5 to 10 numbered questions to fill the gaps, which you can answer in shorthand such as "1: yes, 2: see #channel". Stage two builds the document section by section: Claude sets up the full structure, starts with the section that has the most unknowns, asks questions for each section, brainstorms 5 to 20 possible points for you to keep, remove or combine, drafts the section and makes targeted edits based on your feedback, leaving summary sections for last. Stage three is reader testing: Claude predicts the questions readers will ask, then hands the document to a fresh Claude with no context from your conversation to see whether it answers correctly and where it finds ambiguity or contradictions. In environments with sub-agents, such as Claude Code, this runs automatically; on the claude.ai website you open a new conversation and test it yourself. What sets it apart is that "will other people understand this?" becomes the final check, which SKILL.md says is there to catch blind spots, things that make sense to the authors but might confuse others, before anyone else reads the document.

Who it's for:people writing proposals, PRDs, technical specs or decision docs who are willing to go back and forth with questions until the document makes sense to someone reading it cold. For a change of a sentence or two, the workflow will feel heavy.

Before you start: the README's install route is to add the anthropics/skills plugin marketplace in Claude Code and install the example-skills plugin (12 example Skills at once), and it says these example Skills are already available on Claude.ai paid plans. If your Claude is connected to tools such as Slack, Teams, Google Drive or SharePoint, the Skill suggests pulling context straight from channels and shared documents, which can include internal team messages; for projects it doesn't recognize, SKILL.md says to ask you before searching. During reader testing the whole document is handed to another Claude instance, so think it through first if the content is sensitive. The Skill is a single SKILL.md with no scripts, and it does not send or publish documents for you. You still need to check facts, links and technical details yourself, and SKILL.md ends by reminding you that you own the document and are responsible for its quality. This folder declares no license, the repository root has no license file, and the README only says that "many" Skills in the repository are open source under Apache 2.0; confirm the terms yourself before copying, adapting or passing on this file. The README says the Skills in this repository are for demonstration and educational purposes and that what you get in Claude may differ from what the repository shows.

Based on the project's public documentation. Not tested by AI Inspiration World.

Source

https://github.com/anthropics/skills/tree/main/skills/doc-coauthoring ↗

Repository ↗

How rankings work →
AI Inspiration World logo: aiw. picked by humansAI Inspiration World logo: aiw. picked by humansAI Inspiration World

Check in daily. The newest in AI lives here.

Updated every day · Latest arrival: Oct 14, 2026

Site version: Oct 11, 2026

About·Methodology·Trust & Safety·@AIInspoWorld