i-have-adhd

Shapes AI responses into action-first, numbered output for readers with ADHD.

9|Updated Jul 17, 2026
One-click install
npx skills add https://github.com/Yassimba/loom --skill i-have-adhd-yassimba
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: i-have-adhd
Source: https://github.com/Yassimba/loom/tree/main/skills/i-have-adhd
Command: npx skills add https://github.com/Yassimba/loom --skill i-have-adhd-yassimba

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Readers with ADHD lose track of multi-step instructions, miss buried next actions, and stall on vague guidance. This Skill restructures every AI response so the next action leads, state is restated each turn, and tangents are suppressed. ## Core Features & Use Cases - Action-first formatting: Leads every response with a concrete next action, numbers multi-step work, and ends with one doable next step. - Persistent session rules: Applies ten output rules (state restating, specific time estimates, visible wins, capped lists, no preamble or closers) for the whole session until the user says "stop adhd mode". - Use Case: A developer with ADHD asks an agent to fix a failing auth test. Instead of a long explanation, the response opens with the exact command to run, lists three numbered steps, and ends with one two-minute action. ## Quick Start Use the i-have-adhd skill to make all your responses action-first and easy to execute for the rest of this session.

Frequently Asked Questions about i-have-adhd

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make AI responses easier to follow with ADHD?

Invoke the i-have-adhd skill to switch the assistant into an action-first output mode. It leads with the next action, numbers multi-step tasks, restates progress each turn, and removes preamble and closing filler for the rest of the session.

How do I turn off ADHD output mode once it is active?

Say "stop adhd mode" or "normal mode" and the assistant confirms in one line, then returns to its default response style. The rules otherwise persist for the entire session and across topic changes.

Can I enable ADHD-friendly responses permanently in a project?

Yes. Run loom init --adhd in a project to write the rules into its managed AGENTS.md file. Future loom sync runs keep that section current, and it applies to every agent session in that repository.

When does the skill break its own brevity rules?

It overrides defaults when you ask for a full explanation, when a destructive action needs confirmation, during debug spirals, when a request is genuinely ambiguous, or when a harness system prompt requires different behavior.

Does this skill work with agents other than Pi?

Yes. It is a plain SKILL.md instruction set that any agent reading an Agent Skills folder can load, including Claude Code, Codex, OpenCode, Cursor, and Grok. Loom only adds the optional permanent-mode setup.