i-have-adhd Skill: Shorter, More Direct Answers From Your Coding Agent

What ayghri/i-have-adhd does, the 10 output rules it enforces, the before/after it uses as its own example, and how to install it in Claude Code, Antigravity or DeepSeek Harness — no medical claims, just an output style.

Last updated: 2026-11-17

Quick answer

i-have-adhd is an output style for coding agents, not a productivity app and not a health tool. Its repository summary: "A skill to stop your coding agent from burying the answer. ADHD-friendly output." The README banner adds the part that matters for how you describe it: "ADHD-friendly outputs. No ADHD diagnosis needed!"

The reference implementation is ayghri/i-have-adhd, MIT licensed. At the GitHub API snapshot of 2026-11-17 it had 61,410 stars and 3,603 forks, created 2026-05-13 and last pushed 2026-10-19. The README ships translations in ten languages under .github/readme/, and credits the book the rules were loosely adapted from: The Adult ADHD Tool Kit by J. Russell Ramsay and Anthony L. Rostain — with the caveat, in the repo's own words, that they are "Adapted for how an LLM should respond, not how a human should organize their day."

That distinction is the whole point of this page: it changes the shape of a reply. It does not say anything about the reader's brain.

The before/after it uses itself

The README's clearest explanation is a side-by-side. Before, the agent writes a paragraph like this:

Great question! Let me think about this. Your auth flow has a few moving pieces: the middleware, the token verification, and the cookie handling. Looking at src/auth.ts, the verifyToken function (around lines 42-58) seems to be using an older jsonwebtoken API. One approach would be to update the package and rewrite that function. After making the change, you'd want to run the auth tests to confirm nothing breaks. By the way, you might also want to look at your dependency versions overall. Hope this helps! Let me know if you want to dig deeper.

After:

Edit src/auth.ts:42 to update the token validation.

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

Next: paste the first failing line if any test fails.

Same information, same correctness, roughly a tenth of the reading. If the second one is what you want from an agent by default, the skill is the cheapest way to get it consistently.

The 10 rules

The full text lives in skills/i-have-adhd/SKILL.md; the README lists them as:

#Rule
1Lead with the next action
2Number multi-step tasks
3End with one concrete next step
4Suppress tangents
5Restate state every turn
6Specific time estimates (minutes, not "a bit")
7Make wins visible
8Matter-of-fact errors
9Cap lists to 5 items
10No preamble. No recap. No closers

Rules 4 and 6 are the ones people notice later. Suppressing tangents removes the "by the way, you might also want to…" paragraph that quietly turns a one-line fix into an afternoon; concrete time estimates replace "a bit" with "about 10 minutes", which is either useful or obviously wrong, and both are better than vague.

Install it

The README's primary instruction is deliberately conversational — paste this into your agent's prompt:

Install the i-have-adhd skill/plugin from https://github.com/ayghri/i-have-adhd, refer to the repo's AGENTS.md for instructions.

INSTALL.md then documents per-agent routes. Antigravity is explicit:

agy plugin install https://github.com/ayghri/i-have-adhd
agy plugin list

To remove it later: agy plugin uninstall i-have-adhd, or keep it installed and disable it with agy plugin disable i-have-adhd.

Because the rules are prose, they also work as a plain instruction block. INSTALL.md shows an "always-on" variant for Antigravity that appends the ten rules to ~/.gemini/GEMINI.md — the same block can go into an AGENTS.md for any agent that reads one.

Customising it

The README documents the fork swap, and it is fiddly enough to be worth copying exactly:

claude plugin uninstall i-have-adhd            # drop the upstream copy first:
claude plugin marketplace remove i-have-adhd   # fork and upstream share both names
claude plugin marketplace add <your-username>/i-have-adhd
claude plugin install i-have-adhd@i-have-adhd

Then restart your coding assistant and re-invoke /i-have-adhd. The comment in the README explains why the removal step is not optional: the fork and the upstream copy share both names.

Using it with DeepSeek Harness

A dsh skill is the same object as a Claude Code skill — a folder holding a SKILL.md — so the install is a copy:

git clone https://github.com/ayghri/i-have-adhd
mkdir -p .dsh/skills && cp -r i-have-adhd/skills/i-have-adhd .dsh/skills/

Project skills live at .dsh/skills/<name>/SKILL.md, and dsh re-injects the catalogue on the next prompt, so no restart is needed. The step-by-step version is in installing skills and skill installation, hands-on; if you want the folder layout explained, writing a skill covers it.

One thing to decide before you copy: this skill changes tone for the whole session. If you want it only for long build logs or debugging sessions, keep it as a project skill in that project rather than a global one.

Where it sits next to other skill packs

  • Versus a written style guide. A style guide has to be pasted or remembered. A skill is loaded by the agent, so it applies without you nagging.
  • Versus Jane Street agent skills or pstack. Those are workflow packs — playbooks for planning, review and verification. This one is a single formatting contract. They compose: the workflow decides what to do, this decides how the answer is laid out.
  • Versus a linter. A linter checks code. This checks the message that explains the code.

If you are deciding what to install first, the best dsh skills puts the packs side by side, and andrej-karpathy skills is the other common "make the agent behave" install.

Limits worth stating

  • It is a style, not an accuracy fix. Shorter answers are easier to check; they are not more likely to be right. Rule 8 says errors should be stated matter-of-factly, which helps you spot them, and nothing more.
  • "ADHD-friendly" is the project's own label. It comes from a book adaptation, not from a clinical study, and the README does not claim otherwise. If you need accessibility accommodations at work, this skill is not one — it is a formatting preference.
  • Five-item list caps bite. Rule 9 truncates long lists, which is great for a status update and bad for a migration checklist. Fork and edit when a project needs the long version.
  • It overrides your own tone. If your team's convention is a chatty, explanatory reply, this is the wrong default; scope it to a project instead.

FAQ

What is the i-have-adhd skill? An output-style skill for coding agents: "A skill to stop your coding agent from burying the answer. ADHD-friendly output." It rewrites replies — action first, numbered steps, one next step, no preamble, no closer.

Is it a medical tool or a diagnosis? No. The README says "ADHD-friendly outputs. No ADHD diagnosis needed!" and describes the rules as loosely adapted from a book about adult ADHD for how an LLM should respond. It is a readability format, not a health product.

How do I install it? The README's shortest route is to paste "Install the i-have-adhd skill/plugin from https://github.com/ayghri/i-have-adhd, refer to the repo's AGENTS.md for instructions." into your CLI. Antigravity uses agy plugin install https://github.com/ayghri/i-have-adhd.

Can I customise the rules? Yes. Fork it, edit skills/i-have-adhd/SKILL.md, then uninstall the upstream copy, remove the shared marketplace name, and add your fork.

Does it work in DeepSeek Harness? Yes — copy skills/i-have-adhd into .dsh/skills/. There is no separate dsh plugin in the repository.

Sources

Every claim above comes from these pages, read on 2026-11-17:

  • github.com/ayghri/i-have-adhd — README: the repository summary, "ADHD-friendly outputs. No ADHD diagnosis needed!", the before/after pair quoted above, the ten rules, the fork-and-swap commands with their explanations, the book credit line, the YouTube breakdown link, and the MIT licence. GitHub API snapshot taken 2026-11-17: MIT, 61,410 stars, 3,603 forks, created 2026-05-13, pushed 2026-10-19 (a snapshot, not a fixed number).
  • INSTALL.md — the per-agent install sections, including the Antigravity commands (agy plugin install, agy plugin list, agy plugin uninstall, agy plugin disable) and the "always-on" GEMINI.md block.
  • .github/readme/ — the ten translated READMEs linked from the language row.

The project is a community skill, not an Anthropic, Google or DeepSeek product, and it is not affiliated with this site. Commands come from the README and INSTALL.md at the date shown; re-check them before running, because plugin names change with releases.

Related reading: installing skills, skill installation, the best dsh skills, Jane Street agent skills, andrej-karpathy skills.

DSH Plugins is an independent community directory of DeepSeek Harness plugins. Not affiliated with or endorsed by DeepSeek. Third-party plugins are not security-audited — review the source before installing.

New DeepSeek Harness plugins, weekly. No spam.