Agent skills

i-have-adhd

ayghri/i-have-adhd

A ten-rule skill that stops your coding agent burying the answer under preamble.

Repository
Stars
18.8k
Forks
980
Open issues
17
Language
Python
Licence
MIT
Created
2026-05-13
Last push
2026-08-10
Effort to adopt

Plug in

Install and use the same day. Little to no custom code.

01

What it actually does

Coding agents pad. You ask a direct question and get "Great question! Let me think about this..." followed by three paragraphs of context before the actual instruction appears, and a "Hope this helps!" at the end.

i-have-adhd is a skill that enforces ten rules on the output. Lead with the next action. Number multi-step tasks. End with one concrete next step. Suppress tangents. Restate state every turn. Give specific time estimates in minutes, not "a bit". Make wins visible. Report errors matter-of-factly. Cap lists at five items. No preamble.

The README's before-and-after is the clearest possible pitch: a rambling paragraph about your auth flow becomes "Run `npm install jsonwebtoken@latest`, then edit `src/auth.ts:42`" followed by three numbered steps.

Despite the name, no diagnosis is required or implied — it is an output-formatting skill, and it is translated into six languages.

02

Who it's for

  • 01

    Anyone who reads agent output on a phone or between meetings and needs the answer first

  • 02

    Teams standardising how agents report status so updates are scannable

  • 03

    Engineering managers drowning in verbose agent-generated summaries

  • 04

    People who find long preamble genuinely hard to parse, for whatever reason

EngineeringOperationsFoundersAny industry
03

Where it earns its keep

  • Making agent status updates in Slack short enough that people actually read them
  • Getting a straight answer during an incident instead of a lecture
  • Standardising the shape of automated reports across a team
  • Cutting output tokens — less padding is also less spend
04

Use it, or skip it

Reach for it when

  • Your agent's output is technically correct and practically unreadable
  • Updates are consumed on mobile or in a busy channel
  • You want consistent formatting across everyone's agent sessions
  • You are paying for output tokens spent on pleasantries

Skip it when

  • You genuinely want the reasoning — this compresses explanation, and for teaching or design review that is a loss
  • The audience is a customer, where terse can read as curt
  • You are debugging something subtle and need to see the model's full chain of thought
  • Your team's culture expects narrative context in written updates
05

10 automations

  1. 01Operations

    Scannable Slack status updates

    Apply the skill to every automated agent update posted to Slack so each one leads with the action and caps at five bullets.

  2. 02Engineering

    Incident response formatting

    During an incident, force agent output into numbered steps with time estimates so the responder can act without parsing prose.

  3. 03Founders

    Executive-readable summaries

    Run weekly project reports through the ruleset so the founder gets state, next action, and blockers in under ten lines.

  4. 04Finance

    Output token reduction

    Measure token spend before and after applying the skill across your agent fleet and report the monthly saving.

  5. 05Engineering

    Standardised code review comments

    Make review bots state the required change first and drop the hedging, so authors know exactly what to fix.

  6. 06Sales

    Sales call follow-up notes

    Format post-call summaries as action-first bullets a rep can paste straight into the CRM without editing.

  7. 07Support

    Support response drafts

    Draft customer replies with the fix first, then the explanation — then have a human add warmth before sending.

  8. 08Operations

    Daily standup digest

    Compress overnight agent activity into a five-item maximum digest with a single concrete next step.

  9. 09Marketing

    Marketing brief compression

    Turn sprawling campaign briefs into numbered, time-estimated task lists the content team can execute against.

  10. 10Operations

    Onboarding instructions

    Generate setup guides that lead with the command to run, with realistic minute estimates per step.

Want one of these running by Friday?

LimeDock builds these as real workflows inside your stack — deployed to your cloud, wired into your Slack and CRM, with the code in your repo. You pay a build fee and your own API keys, nothing else.

Book a workflow call
07

Source

Repository stats were read from the GitHub API and reflect the last time we refreshed this entry. The editorial breakdown above is LimeDock’s own analysis — we are not affiliated with ayghri.

https://github.com/ayghri/i-have-adhd