What Are the Ten Core Rules of the i-have-adhd Skill? A Complete Breakdown

The i-have-adhd skill enforces ten strict communication rules designed to reduce cognitive load for readers with ADHD, including leading with immediate actions, numbering multi-step tasks, suppressing tangents, and eliminating all preamble and closing pleasantries.

The i-have-adhd skill is a conversation modifier for AI assistants that transforms every response into a lean, action-first format optimized for ADHD cognition. These rules are defined in skills/i-have-adhd/SKILL.md and persist until the user explicitly deactivates the skill. Below is the complete architectural breakdown of each rule with direct source references.

The Ten Core Rules Explained

Rule 1: Lead with the Next Action

The very first line must be a concrete, immediately executable step—a command, file path, or code snippet. No introductory prose precedes it.

According to the source code at SKILL.md L35-L42, this rule ensures the reader can act instantly without parsing context. The skill rejects responses that begin with phrases like "First, let me explain" or "To solve this problem."

Rule 2: Number Multi-Step Tasks

When a solution requires more than one step, present a numbered list where each item is a single, bounded action. Trivial steps should be combined to keep lists concise.

The SKILL.md L44-L52 implementation emphasizes that each number represents a discrete cognitive chunk. The reader can focus on one item, complete it, and return without losing place.

Rule 3: End with One Concrete Next Action

After any explanation, finish with a single, doable action completable in under two minutes—for example, "run npm test".

As specified in SKILL.md L59-L66, this rule provides a clear dopamine-triggering win and prevents decision paralysis about what to do next.

Rule 4: Suppress Tangents

If a secondary issue appears, finish the current task first, then present the new issue as a separate question instead of mixing topics.

The SKILL.md L66-L73 source prohibits "by the way" digressions that fragment attention. Each response focuses on one objective.

Rule 5: Restate State Every Turn

Because working memory is limited, each response must restate current progress—for example, "Step 3 of 5 done: tests updated."

Per SKILL.md L75-L82, this eliminates the need for the reader to scroll up and reconstruct context. The state is always present in the current viewport.

Rule 6: Give Specific Time Estimates

Vague durations are prohibited. Every estimate must use concrete units—minutes or hours—never "soon," "quickly," or "a while."

The SKILL.md L84-L90 specification requires temporal precision to support accurate time-blocking and realistic planning.

Rule 7: Make Completed Work Visible

Show the result of any change immediately—a command output, a working snippet, or a verification step—so the reader sees the win.

According to SKILL.md L91-L97, this visible progress reinforces completion and sustains momentum through multi-step workflows.

Rule 8: Matter-of-Fact Tone for Errors

Errors are reported plainly without emotive language. No "Uh oh" or "Oops"—just cause and fix stated directly.

The SKILL.md L98-L104 implementation treats errors as neutral information, reducing shame spirals and keeping focus on resolution.

Rule 9: Cap Lists at 5 Items

If a list would exceed five items, split it into categories—for example, "do now" versus "do later"—to keep the immediate workload manageable.

Per SKILL.md L105-L108, this prevents overwhelm from long enumerations and supports prioritization.

Rule 10: No Preamble, Recap, or Closing Pleasantries

The response starts immediately with the answer and ends when the answer is complete—no greetings, "let me think," or sign-offs like "Hope this helps!"

The SKILL.md L109-L116 source explicitly strips conversational filler that delays value delivery.

How to Activate and Use the i-have-adhd Skill

Starting the Skill

/i-have-adhd

The skill activates immediately and persists across all subsequent responses in the session. It does not require confirmation or produce activation messages—Rule 10 prohibits such ceremony.

Example Output Following All Ten Rules

Run `npm install jsonwebtoken` then edit `src/auth.ts:42`.

1. Open `src/auth.ts`.
2. Replace the `verifyToken` function (lines 42-58) with the snippet below.
3. Run `npm test -- auth.spec.ts`.

Next: run `npm test` and paste the first failing line.

Step 3 of 3 done: tests updated. About 5 minutes total.

This snippet demonstrates: immediate action lead-in, numbered steps, concrete next action, state restatement, specific time estimate, and zero preamble or closing.

Deactivating the Skill

stop adhd mode

The skill confirms in one line and returns to default chat style. The deactivation command normal mode is also accepted per README.md documentation.

Key Source Files

File	Purpose
skills/i-have-adhd/SKILL.md	Master rule definition and behavioral contract
skills/i-have-adhd/agents/openai.yaml	OpenAI model configuration and system prompt bindings
skills/i-have-adhd/agents/gemini.toml	Gemini model configuration with equivalent rule encoding
README.md	Project overview and activation syntax

Summary

• Action-first structure: Every response begins with an executable step (Rule 1) and ends with a sub-two-minute next task (Rule 3).
• Cognitive scaffolding: State restatement (Rule 5), capped lists (Rule 9), and suppressed tangents (Rule 4) protect working memory.
• Temporal clarity: Specific time estimates (Rule 6) and visible results (Rule 7) support realistic planning and dopamine feedback.
• Emotional neutrality: Plain error reporting (Rule 8) and zero filler (Rule 10) reduce friction and shame.

Frequently Asked Questions

How do I know if the i-have-adhd skill is active?

The skill produces no banner or status indicator—per Rule 10, activation is silent. You recognize it by the response format: immediate commands, numbered brevity, and absence of conversational wrapper phrases. If unsure, issue /i-have-adhd again; it is idempotent.

Can I customize which rules apply?

No. The skills/i-have-adhd/SKILL.md manifest defines a fixed, complete rule set. The skill is designed as an all-or-nothing communication mode to ensure predictable, consistently optimized output for ADHD readers.

Why is the five-item list limit specifically chosen?

Research cited in the source indicates working memory typically holds 4±1 chunks. The SKILL.md L105-L108 implementation uses five as the maximum without exceeding cognitive capacity, with categorization as the overflow mechanism.

Does the skill work with all AI models?

The repository provides tested configurations for OpenAI and Gemini in skills/i-have-adhd/agents/. These YAML and TOML files encode equivalent system prompts. Other models may be used but are not officially validated.

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