AI Briefing
KO

Paperthin Releases 28 Skills for Coding Agents

·2026.09.15 21:00

Key point

Paperthin, a collection of 28 open-source skills designed to resolve discrepancies between documentation and code generated by coding agents, has been released.

1 / 4

Details

Paperthin, an open-source collection of skills designed to resolve discrepancies between documentation and code when using coding agents, has been released. The project draws inspiration from Matt Pocock's 'Skills For Real Engineers' structure and philosophy but is designed as an independent workflow rather than a simple fork.

Key Features and Classification System

Paperthin consists of 28 SKILL.md files, classified into categories such as depth, breadth, coil, and mesh based on artifact count and time criteria. Among these, 19 skills belong to the depth category, which handles the accuracy, rewriting, and compression of single artifacts. Representative skills include re0, which cleanly rewrites artifacts to a v0 state, factchk, which performs fact-checking and error removal, and dedash, which eliminates duplicates.

Installation and Usage

Users can install the package using the command npx skills@latest add LilMGenius/paperthin --global --agent '*'. By default, skills can be invoked by both the model and the user, but setting disable-model-invocation: true in the metadata restricts invocation to users only. The license is MIT, allowing free personal and commercial use.

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.