# What Conversational Elements Are Forbidden by Rule 10 in the i-have-adhd Skill?

> Discover the three conversational elements forbidden by Rule 10 in the i-have-adhd skill: preambles, recaps, and closers. Learn how this rule ensures immediate actionable information.

- Repository: [Ayoub Ghriss/i-have-adhd](https://github.com/ayghri/i-have-adhd)
- Tags: deep-dive
- Published: 2026-08-27

---

**Rule 10 forbids three types of conversational elements: preambles/openers, recaps, and closers—ensuring every response begins and ends with actionable information.**

The `ayghri/i-have-adhd` repository defines a specialized skill for supporting ADHD users through structured, distraction-free conversational patterns. Rule 10—internally labeled as **Rule IO** for "No preamble, no recap, no closing pleasantries"—imposes strict constraints on language generation to maintain laser focus on immediate next steps.

## The Three Forbidden Conversational Elements

Rule IO explicitly bans three categories of language that add cognitive overhead without advancing action.

### Openers and Preambles

Any introductory phrase that delays the actionable content violates this rule.

| Forbidden pattern | Why it fails |
|---|---|
| "Great question" | Acknowledgment without action |
| "Let me think about this" | Self-reference that postpones the answer |
| "Sure!" | Empty affirmation |
| "Looking at your..." | Setup language before the actual instruction |

**Correct alternative:** Start immediately with the command—`Run `npm install jsonwebtoken`, then edit `src/auth.ts:42`.`**

### Recaps and Mid-Message Summaries

Restating completed work or requesting unnecessary follow-up adds redundancy that disrupts task momentum.

| Forbidden pattern | Why it fails |
|---|---|
| "I've now done X, Y, and Z, which means..." | Repeats information the user already processed |
| "Hope this helps" | Unverifiable sentiment with no next step |
| "Let me know if you need anything else" | Opens a loop when the interaction is complete |

### Closers and Polite Sign-Offs

Farewell-style language creates artificial conversation endpoints that don't advance the workflow.

| Forbidden pattern | Why it fails |
|---|---|
| "Happy to clarify" | Offers future help without defining trigger |
| "Feel free to ask" | Vague invitation without concrete condition |
| "Hope this helps" | (Also banned as recap—appears in both categories) |

## Source Code Location and Implementation

The Rule IO specification resides in [`skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/skills/i-have-adhd/SKILL.md) at **lines 107–113**. According to the source, the rule ensures **"the first line is always the actionable command and the last line is the next concrete step, without any fluff."**

Related documentation files include:

- **[`skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/skills/i-have-adhd/SKILL.md)** – Canonical rule definitions (including all ten formatting rules)
- **[`README.md`](https://github.com/ayghri/i-have-adhd/blob/main/README.md)** – Skill overview with summarized rules list
- **[`AGENTS.md`](https://github.com/ayghri/i-have-adhd/blob/main/AGENTS.md)** – Runtime integration patterns showing where Rule IO enforcement occurs

## Correct vs. Incorrect Examples

### Compliant Response (Rule IO Followed)

```text
Run `npm install jsonwebtoken`, then edit `src/auth.ts:42`.
Next: run `npm test` and paste the first failing line.

```

Structure analysis:
- Line 1: Immediate command with specific file path
- Line 2: Explicit next step with continuation trigger

### Violating Response (Rule IO Broken)

```text
Great question! Let me think about this.
Here's the fix. Next: run `npm test`.
Hope this helps.

```

Violation breakdown:
1. "Great question! Let me think about this." → **Preamble**
2. "Hope this helps." → **Closer**

## Why These Constraints Matter for ADHD Support

The `i-have-adhd` skill design prioritizes **cognitive load reduction** through structural predictability. By eliminating variable-length introductions and indeterminate endings, each response becomes:

- **Scannable** – Users know the first line contains the action
- **Bounded** – Clear endpoint prevents open-loop anxiety
- **Actionable** – Every word either instructs or defines the next trigger

This pattern aligns with executive function research: predictable structure reduces decision fatigue and supports task initiation.

## Summary

- **Rule IO (Rule 10)** forbids three elements: preambles, recaps, and closers
- **Source location:** [`skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/skills/i-have-adhd/SKILL.md), lines 107–113
- **Core mandate:** First line = actionable command; last line = next concrete step
- **Design goal:** Eliminate cognitive overhead for ADHD users through rigid conversational structure

## Frequently Asked Questions

### What does Rule 10 stand for in the i-have-adhd skill?

Rule 10 is internally labeled **Rule IO**, representing "No preamble, no recap, no closing pleasantries." The "IO" mnemonic emphasizes input-output efficiency: minimal words between user need and system response.

### Why are polite phrases like "Hope this helps" banned?

These phrases create **false conversation endpoints** without advancing the task. For ADHD users, they introduce ambiguity about whether action is required, potentially triggering rumination or avoidance. The source code explicitly categorizes "Hope this helps" as both a recap and closer violation.

### Where are the i-have-adhd rules enforced?

Rule IO is defined in [`skills/i-have-adhd/SKILL.md`](https://github.com/ayghri/i-have-adhd/blob/main/skills/i-have-adhd/SKILL.md) and implemented across runtime integrations documented in [`AGENTS.md`](https://github.com/ayghri/i-have-adhd/blob/main/AGENTS.md). The rules operate as formatting constraints on language model outputs when the skill is active, not as runtime code checks.

### What's the difference between Rule 10 and other rules in the skill?

Rule IO is one of ten formatting rules in the skill definition. While other rules govern structure (like bullet formatting or heading hierarchy), Rule IO specifically constrains **conversational tone**—making it unique in targeting language warmth rather than visual layout.