I Have ADHD

I Have ADHD gives coding agents a response format with the next action up front, fewer tangents and a visible record of progress.

Turn a debug update into one next action

Bring a long status message and identify the command and evidence that must survive the rewrite.

Which debugging update is hard to act on?

Include the current blocker, failing check and any paths or error codes.

Copy the request into a new Floatboat conversation and provide the required material.

Download Floatboat

Your request for Floatboat

Preview and copy the request. It is not sent automatically from this page.

Reference: I Have ADHD
https://github.com/ayghri/i-have-adhd/blob/main/skills/i-have-adhd/SKILL.md

Read the original documentation and check the current environment and dependencies. Propose any new installation or tool connection before proceeding.

My task:
Use I Have ADHD to rewrite this debugging update. Lead with the exact test to rerun. Keep the failing assertion, error code and current blocker. If one detail is uncertain, say which check resolves it.

Expected deliverable:
An action-first update with the exact check, state, and remaining decision.

What I Have ADHD does

The Skill is an instruction file for response structure. It puts the next action first, breaks work into bounded steps and keeps the current state visible across turns. It preserves extra detail when the reader asks for an explanation or when ambiguity and safety make it necessary. That can trim output, while actual token usage depends on the task and host.

I Have ADHD use cases

Lead with a command when the next action is a check, or a decision when the reader must choose. Keep the error evidence and approval conditions that make either action safe.

Unblock a failing test

Start with the test command and name the first failing assertion. Put the longer explanation after the action so a reader can reproduce the problem without scanning the entire update.

Pick up interrupted work

Write the current checkpoint, completed check and remaining action in a small status update. When you return, you can continue from the same file and revision without asking where work stopped.

Choose between two implementations

Lead with the decision to make, then state the trade-off and consequence for each option. Keep the dependency or migration detail that could change the answer.

Ask before a destructive action

Put the requested approval and affected target first. Preserve the consequences and recovery path so a shorter reply still supports an informed decision.

How to use I Have ADHD

Try the instruction on a task you can verify. Compare the opening line, the actual next action and whether the shorter answer still includes the evidence you need.

Load the response Skill
Use the upstream install guide for your host and invoke the Skill in a coding-agent session.
Specify the useful details
Give it one task and name information that must survive the rewrite, such as paths, commands or approval conditions.
Check the next action
Read the first line and confirm it can be acted on. Ask for more detail if evidence or the reason for a decision is missing.

Explore More Agent Skills

Read another Skill guide and check its original source before using it.

I Have ADHD FAQ

What does the Skill change?

It changes how the agent presents an answer: an action comes first, and multi-step work is broken into short, ordered actions.

How do I tell whether a shorter reply saves tokens?

Compare output usage on the same task and model, then check that both replies preserve the required commands and evidence. The Skill does not alter provider limits or billing rules.

Can it still explain a difficult topic?

Yes. The source allows fuller explanations when requested while keeping the structure easy to scan.

Does it change input context or tool logs?

The instruction primarily shapes the response. Tool-output indexing or retrieval requires a separate workflow.

Will it stay active in a new session?

The rules apply within a session. Reload the Skill or verify a documented host-specific startup option for new sessions.

Should commands and warnings be abbreviated?

Preserve exact commands, paths and safety conditions. The upstream treats risky or ambiguous actions as exceptions to brevity.