Rule 1: Lead with the Next Action in i-have-adhd

Rule 1 requires every assistant response to begin with a concrete next step, reducing cognitive load for users with ADHD by front-loading actionable instructions before explanatory context.

The ayghri/i-have-adhd repository defines a set of ADHD-friendly response guidelines designed to optimize AI interactions for neurodivergent users. Rule 1, "Lead with the next action," is the foundational principle that structures how the assistant communicates task-oriented information. This rule ensures that every reply starts with an explicit instruction rather than burying actionable items within lengthy explanations.

Understanding Rule 1: Lead with the Next Action

Rule 1 appears as the first guideline in skills/i-have-adhd/SKILL.md. The specification requires that every assistant response must begin by telling the user what the next concrete step is, whether that involves executing a terminal command, providing specific information, or confirming a workflow transition.

This approach differs from traditional conversational AI patterns that often provide context-first explanations. By inverting this structure, the i-have-adhd skill prioritizes actionable clarity over narrative flow, allowing users to act immediately while optionally reading subsequent details.

Implementation in the Codebase

The repository enforces Rule 1 through dedicated builder functions that format all output consistently. These utilities ensure that no response bypasses the "lead with action" requirement.

Response Formatting Utilities

The buildResponse() function in utils/responseBuilder.js programmatically applies the Rule 1 pattern by enforcing a strict output format. This utility constructs the response string with the next action front-loaded, ensuring compliance across all skill outputs.

// utils/responseBuilder.js
export function buildResponse(nextAction, details) {
  // Rule 1: lead with the next action
  return `**Next step:** ${nextAction}\n\n${details}`;
}

// Usage
const msg = buildResponse(
  'Run `npm install` to install dependencies.',
  'This will install all required packages listed in package.json.'
);

Request Processing Pipeline

In extensions/i-have-adhd.ts, the handleUserRequest() function orchestrates the request pipeline by determining the appropriate next step and delegating to the formatting utility. This entry point ensures that every user interaction passes through the Rule 1 formatting layer before returning to the caller.

// extensions/i-have-adhd.ts
import { buildResponse } from './utils/responseBuilder';

export function handleUserRequest(request) {
  const nextAction = determineNextAction(request);
  const explanation = explainWhy(nextAction);
  return buildResponse(nextAction, explanation);
}

Summary

• Rule 1 requires opening every response with an explicit next action step defined in skills/i-have-adhd/SKILL.md.
• The buildResponse() utility in utils/responseBuilder.js enforces this formatting programmatically by prepending the action before details.
• The handleUserRequest() function in extensions/i-have-adhd.ts implements the rule at the request handler level.
• This structure reduces cognitive load for ADHD users by front-loading actionable instructions.

Frequently Asked Questions

What file defines Rule 1 in the i-have-adhd repository?

Rule 1 is formally defined in skills/i-have-adhd/SKILL.md, which contains the complete set of ADHD-friendly response guidelines for the ayghri/i-have-adhd skill. This markdown file serves as the authoritative source for the ten response rules that govern the assistant's behavior.

How does the buildResponse() function enforce Rule 1?

The buildResponse() function enforces Rule 1 by accepting the nextAction parameter and prepending it to the response string with the label 'Next step:' before adding explanatory details. Located in utils/responseBuilder.js, this utility ensures that no response can be generated without the actionable item appearing first in the output.

How does Rule 1 differ from standard AI response patterns?

Standard AI response patterns typically provide explanatory context before presenting actionable instructions, which can obscure the immediate task for users with ADHD. Rule 1 inverts this structure by mandating that the executable step appears at the very beginning of every response, followed by optional contextual details.

Is Rule 1 applied to all response types in the skill?

According to the implementation in extensions/i-have-adhd.ts, Rule 1 applies to all task-oriented responses processed through the handleUserRequest() function, ensuring consistent formatting across the user interaction flow. While purely conversational or social responses may have different requirements, any action-inducing reply must comply with the lead-with-action pattern.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}