'i-have-adhd' Skill Released to Improve Readability of AI Coding Agent Responses
Key point
An open-source skill has been released to prevent verbose responses from AI coding agents by prioritizing core action instructions and step-by-step guidance.
Details
The open-source project 'i-have-adhd', which has received 28.8k stars on GitHub, has been released. This skill is designed to be applied to major AI coding agent tools such as Claude, Codex, Cursor, and OpenCode to prevent responses from becoming verbose or burying key information.
Core Features and Rules This skill focuses on reducing developers' cognitive load by applying an ADHD-friendly output format. The main rules are as follows:
- Action First: Presents the next action instruction as the top priority over explanations.
- Step-by-Step Numbering: Separates multi-step tasks with clear numbering.
- Concrete Conclusion: Presents specific next steps instead of closing remarks like 'Hope this helps!'.
- Maintain Conciseness: Suppresses introductions, summaries, and tangents, and limits lists to a maximum of 5 items.
Installation and Customization
Users can easily apply it by entering the installation command in the CLI prompt or referring to INSTALL.md. Additionally, rules can be customized to individual preferences by modifying the skills/i-have-adhd/SKILL.md file. This project follows the MIT license and is characterized by adapting adult ADHD toolkit theory to LLM response styles.
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.