ADHD-Crafted Claude Skill Forces Actionable, Tangent-Free Outputs
Phelps1576 · reddit · 2026-07-28
A Reddit user shared a custom Claude Skill (prompt template) called i-have-adhd, designed to shape model outputs specifically for readers with ADHD.
The skill forces Claude to abandon verbose, rambling responses in favor of strict rules:
- Lead with action: The first line must be an immediate, concrete action (like a command), not context.
- Number multi-step tasks: Break down work into bounded, numbered lists without chaining "and then."
- Externalize state: Restate progress every turn (e.g., "Step 3 of 5 done") to compensate for small working memory.
- Suppress tangents: Ban pleasantries and recaps; state errors matter-of-factly with direct fixes.
- Specific time estimates: Provide concrete time units (e.g., "About 15 minutes") instead of vague descriptions.
More from coding & agent
- Tokyo Codex meetup on Aug. 7 will center on CLI contributions and agent workflows — paw_lean · 2026-07-28
- Open-source benchmark tracks coding-agent token usage without leaking prompts — bestofdesp · 2026-07-28
- Matt Pocock says his AGENTS.md pruning lesson became the longest in the course — mattpocockuk · 2026-07-28
- A stronger model can still hurt your AI product, so test upgrades on production tasks first — max_gladysh · 2026-07-28
- Claude Code says terminal-first agents beat the copy-paste ChatGPT workflow — Aiden_Tech_Ai · 2026-07-28
- Turn Claude Code into a design pipeline instead of a one-shot prompt — PrajwalTomar_ · 2026-07-28