I Have ADHD Skill
Stop your coding agent from burying the answer: action first, numbered steps, no fluff.
- Source repo
- ayghri/i-have-adhd
- Stars
- ★ 58k
- Last updated
- 7d ago
- License
- MIT
- Primary language
- Python
- FA score
- 46/100 · Major gaps
At a glance
- How it runs
- Works with
- Platform-specificClaude Code
- Cost
- Free, no paid service needed
- Setup effort
- Medium · a few setup steps
- You'll need
- Typical use
- A developer using Claude Code wants quick answers without long-winded explanations.
- Not a fit if
- Users of Cursor, Codex, or other coding assistants
- Non-technical users unwilling to edit SKILL.md
- Teams wanting a standalone agent, not a style plugin
- Source review
- 46/100 · Major gaps
What does this agent do, and when should you use it?
A skill for Claude Code that enforces an ADHD-friendly output style: lead with the next action, number multi-step tasks, and cut all preamble. It's not a standalone agent but a response-style modification installed as a Claude plugin. Invoke with /i-have-adhd after installation. Based on The Adult ADHD Tool Kit, adapted for LLM responses. MIT licensed, with multi-language READMEs.
The skill rewrites Claude Code's response behavior to follow 10 rules: lead with next action, number multi-step tasks, end with a concrete next step, suppress tangents, restate state every turn, give specific time estimates (minutes not 'a bit'), make wins visible, matter-of-fact errors, cap lists at 5 items, and no preamble/recap/closers. It achieves this via a SKILL.md file and is managed through the Claude plugin marketplace.
- A developer using Claude Code wants quick answers without long-winded explanations.
- Someone with ADHD or short attention span needs concise, structured instructions.
- A programmer debugging wants the assistant to point directly to the fix location and steps.
- A user wants to cut through polite niceties and get to the point for faster work.
- A team wants a consistent, no-nonsense output style across all Claude Code sessions.
How do you install or deploy this agent?
To be added. The README mentions installation instructions in INSTALL.md, but they are not provided here. Refer to the repository's INSTALL.md file.
How do you use this agent?
After installation, invoke the skill in Claude Code with the command /i-have-adhd. The skill then adjusts the assistant's response style.
What are this agent's strengths and limitations?
- Directly provides actionable steps, saving reading time.
- Based on an evidence-based ADHD toolkit, giving credibility.
- MIT license permits free use and modification.
- Multi-language support (English, Simplified Chinese, Portuguese, Japanese, Vietnamese, Korean).
- Only works with Claude Code, not other coding assistants.
- Installation may be complex, requiring manual plugin marketplace commands.
- Customization requires editing SKILL.md, needing technical know-how.
- Not a standalone agent; depends on the Claude Code environment.
How does this agent compare with similar options?
Key facts side by side with the most closely related agents.
| Agent | Source review | Form / cost | Stars | Updated | Language | Full support on |
|---|---|---|---|---|---|---|
| I Have ADHD Skill This agent | 46 · Major gaps | Agent plugin / skillFree | ★ 58k | 7d ago | Python | Claude Code |
| Fable Method – Distilled Agent Workflow | 55 · Major gaps | Agent plugin / skillFree + model costs | ★ 2.3k | 16d ago | Python | Claude Code |
| Ponytail | 50 · Major gaps | Agent plugin / skillFree + model costs | ★ 167k | 1d ago | JavaScript | Codex · Claude Code |
| Continue | 43 · Major gaps | CLIFree + model costs | ★ 36k | 3mo ago | TypeScript | — |
How does FollowAgents rate this agent?
Why each dimension lost points
Evidence: The repository includes a GitHub Actions workflow with least-privilege permissions (contents: read, pull-requests: read, issues: read, id-token: write, actions: read), but no user confirmation mechanism is found. Data flow transparency is limited; no explicit statement on how data is processed or transmitted. Sensitive data handling is not mentioned. Dependency security is not assessed, but dependencies are minimal. External effects include modifying user config directory (CLAUDE_CONFIG_DIR) and writing flag files, but consent is not explicit. Rollback mechanism is not mentioned. Source attribution: author is Ayoub Ghriss, but not verified. Deductions: lack of user confirmation, opaque data flow, no sensitive data handling, dependency security not assessed, external effects without explicit consent, no rollback, unverified source.
Evidence: The repository shows good internal consistency; README, SKILL.md, and test files correspond. Dependency availability: dependencies are minimal, but runtime dependencies are not explicitly listed. Failure messages: tests include error messages, but no user-friendly failure prompts. Deductions: dependency availability not explicit, failure messages insufficient.
Evidence: Target audience is clear (ADHD readers), scenarios include coding assistant outputs. Capability boundaries: limitations of the skill are not explicitly stated. Trigger precision: triggered via commands or hooks, but not detailed. Environment fit: supports multiple runtimes (node, sh, powershell), but not all environments are specified. Deductions: capability boundaries unclear, trigger precision and environment fit could be improved.
Evidence: Information architecture is clear, with README, INSTALL.md, SKILL.md, etc. Install notes exist. Naming stability: names are consistent. Examples and FAQ: examples exist, but no FAQ. Known limitations: not explicitly listed. License: MIT. Versioning and changelog: version number exists, but no changelog. Maintenance responsibility: not explicit. Deductions: missing known limitations, changelog, and maintenance responsibility.
Evidence: Output usability: provides concrete output format examples. Marginal value: unique value for ADHD-specific needs. Cost-benefit: no performance or cost data provided. Deductions: cost-benefit not assessed.
Evidence: Claim traceability: claims in README partially correspond to code. Cross-source corroboration: test files support some claims. Fact-inference separation: not explicitly distinguished. Deductions: claim traceability incomplete, cross-source corroboration limited, fact-inference not separated.
- The skill modifies the user's CLAUDE_CONFIG_DIR and writes flag files, which may affect the user's environment without explicit consent.
- Dependency security is not assessed; it is recommended to check all dependencies for known vulnerabilities.
- Publisher identity is unverified; use with caution.
FAQ
What are the prerequisites for installation?
How do I uninstall this skill?
claude plugin uninstall i-have-adhd.