# The 5 Cognitive Principles Behind the i-have-adhd Skill: How ADHD Shapes Information Processing > Discover the 5 cognitive principles of the i-have-adhd skill: limited working memory, knowing-doing gap, startup friction, time perception, and dopamine scarcity. Learn how ADHD shapes information processing for actionable AI. - Repository: [Ayoub Ghriss/i-have-adhd](https://github.com/ayghri/i-have-adhd) - Tags: deep-dive - Published: 2026-08-25 --- **The i-have-adhd skill codifies five cognitive facts about ADHD—limited working memory, the knowing-doing gap, startup friction, distorted time perception, and dopamine scarcity—to structure AI responses that are immediately actionable, time-bounded, and visibly rewarding.** The `i-have-adhd` skill in the `ayghri/i-have-adhd` repository transforms how language models communicate with users who have ADHD. By grounding every output rule in documented cognitive science, the skill ensures that technical instructions align with how an ADHD mind actually processes information, rather than forcing users to adapt to conventional, verbose AI formatting. ## The Five Cognitive Facts Defined in SKILL.md According to the canonical definition in [`skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/skills/i-have-adhd/SKILL.md) (lines 23-30), the skill rests on five explicit cognitive facts. These facts directly inform the response-shaping rules that begin at line 31 of the same file. ### 1. Working Memory Is Small Since information not visible on screen is quickly forgotten, the skill never asks users to "keep in mind" context from previous messages. Every response contains all necessary context on-screen, and each step is self-contained. ### 2. Knowing the Answer ≠ Doing the Answer Friction between comprehension and execution causes work to stall. To bridge this gap, every reply begins with a concrete next action the reader can perform immediately, reducing the psychological distance between **comprehension** and **execution**. ### 3. Starting Is the Hardest Step Initiation requires minimal friction. Therefore, the very first line of any response is always the actionable command or code snippet; explanatory prose follows only after the action is presented. ### 4. Time Estimates Feel Uniform Vague durations like "a bit of work" are cognitively processed the same as "a few hours." Consequently, every estimate uses **concrete units** (e.g., "≈ 15 minutes") to enable accurate effort gauging. ### 5. Dopamine Is Scarce Visible progress sustains motivation, while buried wins go unnoticed. The skill highlights completed work immediately ("Login now works with...") and restates each finished step so users see tangible progress throughout the session. ## Implementation Rules in the Source Code These cognitive principles are enforced through strict output rules defined in [`skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/skills/i-have-adhd/SKILL.md). The rules include: - Lead every response with the next physical action - Number multi-step tasks sequentially - Provide precise time estimates for every subtask - Make completed work visible at the start of subsequent responses - Remain active for the entire session unless the user explicitly disables the mode with "stop adhd mode" The skill is declared as a plugin in [`plugin.json`](https://github.com/ayghri/i-have-adhd/blob/main/plugin.json), which enables the `/i-have-adhd` command to activate the mode. A Cursor-compatible mirror exists at [`.cursor/skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/.cursor/skills/i-have-adhd/SKILL.md) to ensure IDE integration. ## Practical Examples of Cognitive-Optimized Responses When activated via the `/i-have-adhd` command, the skill immediately restructures output to align with the five cognitive principles. Single action example: ```text User: /i-have-adhd Assistant: Run `npm install jsonwebtoken`, then edit `src/auth.ts:42`. ``` This demonstrates **leading with the next action**—the first line is a concrete command without preamble—and **keeping context on-screen** by specifying the exact file path. Multi-step task example: ```text User: How do I add OAuth to my Express app? Assistant: 1. Install the library: `npm install passport-oauth2` (≈ 2 minutes). 2. Add the strategy in `src/auth.ts` (≈ 5 minutes). 3. Register the middleware in `app.js` (≈ 3 minutes). Next: Run `npm test` to verify the new endpoint works. ``` Here, the response shows **numbering for limited working memory** (clear sequence), **concrete time estimates** (combating uniform time perception), and a **dopamine-friendly next action** (clear validation step). Each step is bounded and actionable, addressing the knowing-doing gap. ## Summary - The i-have-adhd skill is built on five cognitive facts documented in [`skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/skills/i-have-adhd/SKILL.md) regarding working memory limitations, execution friction, startup difficulty, time perception distortions, and dopamine regulation. - Output rules enforce immediate actionability, precise timeboxing, visible progress tracking, and zero "keep in mind" requirements. - The skill activates via `/i-have-adhd` and persists until "stop adhd mode" is issued, affecting all subsequent model responses in the session. - Implementation spans [`SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/SKILL.md) (definition), [`plugin.json`](https://github.com/ayghri/i-have-adhd/blob/main/plugin.json) (command registration), and [`.cursor/skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/.cursor/skills/i-have-adhd/SKILL.md) (IDE integration). ## Frequently Asked Questions ### What triggers the i-have-adhd skill? The skill activates when a user issues the `/i-have-adhd` command, as declared in [`plugin.json`](https://github.com/ayghri/i-have-adhd/blob/main/plugin.json). Once enabled, it remains active for the entire session, reshaping all AI responses according to the cognitive principles defined in [`skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/skills/i-have-adhd/SKILL.md), until the user explicitly says "stop adhd mode." ### How does this skill differ from standard AI formatting? Standard AI responses often begin with explanatory context or ask users to remember information across turns. The i-have-adhd skill eliminates these cognitive loads by leading with executable actions, providing concrete time estimates, and keeping all necessary context visible on-screen—directly addressing the ADHD-specific challenges of working memory limitations and initiation friction. ### Where are the cognitive principles documented? The five core cognitive facts are explicitly listed in [`skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/skills/i-have-adhd/SKILL.md) at lines 23-30, with the corresponding output-shaping rules beginning at line 31. This file serves as the canonical definition, with a Cursor-specific mirror maintained at [`.cursor/skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/.cursor/skills/i-have-adhd/SKILL.md) for IDE compatibility. ### Can the skill be used outside of Cursor? While the repository includes a Cursor-specific implementation at [`.cursor/skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/.cursor/skills/i-have-adhd/SKILL.md), the core skill definition in [`skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/skills/i-have-adhd/SKILL.md) is runtime-agnostic. Any AI system that supports plugin-based skills and command triggers can implement these ADHD-optimized cognitive principles.