i-have-adhd

Product Introduction

1. Definition: The i-have-adhd is an open-source, behavior-modifying plugin/skill for AI coding assistants like Claude, Cursor, and GitHub Copilot. It functions as a prompt engineering layer that enforces a strict, structured output format on the AI.
2. Core Value Proposition: It exists to eliminate verbose, meandering, and context-burying responses from AI coding agents, delivering direct, actionable, and ADHD-friendly outputs that prioritize developer productivity and reduce cognitive load during software development.

Main Features

1. Structured Output Enforcement: The skill injects a set of 10 concrete rules into the AI's system prompt. How it works: It modifies the assistant's response behavior by mandating a "next action first" approach, numbered steps for multi-part tasks, and the suppression of preambles, recaps, and conversational closers like "Hope this helps!".
2. Multi-Agent & Multi-Platform Compatibility: The repository provides ready-to-use configurations for various AI agent ecosystems. This includes .claude-plugin for Claude Desktop, .cursor/skills/ for Cursor IDE, .agents/plugins for general AI agents, and extensions for other platforms like Gemini, as detailed in the AGENTS.md and INSTALL.md files.
3. Fork-and-Tune Customization: The core rule set is defined in skills/i-have-adhd/SKILL.md. Users can fork the GitHub repository, edit this Markdown file to tailor the rules for their specific workflow, and then install their custom version, allowing for personalized AI interaction protocols.

Problems Solved

1. Pain Point: AI coding assistants often bury crucial code snippets, commands, or answers within lengthy explanations, apologies, or tangential reasoning, forcing developers to waste time scanning and parsing.
2. Target Audience: Software developers, engineers, and technical practitioners who use AI pair programmers, particularly those who value speed, have focus challenges (like ADHD), or work in high-interruption environments. Personas include Full-Stack Developers, DevOps Engineers, and Engineering Managers reviewing AI-generated code.
3. Use Cases: Essential when debugging a complex error where the root cause must be stated immediately; when following a step-by-step migration guide; when extracting a specific code block from a chat history; or during time-pressured tasks like incident response or live coding.

Unique Advantages

1. Strengths & Limitations (Pros & Cons):

   • Pros: Dramatically reduces time-to-answer. Creates predictable, scannable output. Open-source and free (MIT licensed). Lightweight, requiring only a prompt change. Community-vetted with significant traction (58k+ GitHub stars).
   • Cons: Can make AI interactions feel abrupt or less conversational. The rigid structure may occasionally cause the AI to omit useful, nuanced context for extremely complex problems. Effectiveness depends on the underlying AI model's ability to follow the strict instructions consistently.

2. Key Alternatives & Differentiation:

   • Manual Prompt Engineering: Users could manually type "be concise" or "give me just the code" each time. i-have-adhd differentiates by providing a persistent, comprehensive, and finely-tuned rule set that is automatically applied to every interaction, eliminating repetition.
   • Other Output-Formatting Plugins: While other plugins exist for formatting (e.g., for JSON), i-have-adhd is specifically optimized for the cognitive workflow of coding problem-solving, not just syntax. Its rules are based on ADHD management principles, targeting information hierarchy and task initiation.
   • Built-in Agent "Concise" Modes: Some agents have a "concise" toggle. i-have-adhd goes beyond simple conciseness by enforcing a specific action-oriented structure (numbered steps, next step prompting) that actively guides the workflow, rather than just shortening text.

Frequently Asked Questions (FAQ)

1. Do I need an ADHD diagnosis to use the i-have-adhd plugin? No, the i-have-adhd skill is designed for anyone who prefers direct, actionable answers from their AI coding assistant, regardless of whether they have a clinical ADHD diagnosis. It benefits all developers seeking efficiency.
2. How do I install i-have-adhd for Claude or Cursor? Installation is done via the command line or IDE settings. The precise, copy-paste commands for each supported agent (Claude, Cursor, GitHub Copilot, etc.) are provided in the repository's AGENTS.md and INSTALL.md documentation.
3. Can I customize the rules enforced by the i-have-adhd skill? Yes, the i-have-adhd plugin is fully customizable. You can fork the GitHub repository, edit the skills/i-have-adhd/SKILL.md file to modify the 10 core rules, and then install your forked version as a private plugin.
4. Is the i-have-adhd plugin compatible with local LLMs? Compatibility depends on your AI agent framework. If your local LLM client (like LM Studio, Ollama with a frontend) supports a plugin or system prompt injection architecture similar to Claude Desktop, you can likely adapt the skill's rule text for use.
5. What is the difference between a Claude plugin and a Cursor skill in i-have-adhd? They are configuration files tailored to different platforms. The core SKILL.md rules are the same, but the packaging (.claude-plugin folder vs. .cursor/skills/ file) follows the respective platform's specification for integration.