cd /news/ai-agents/better-an-agent-skill-that-asks-for-… · home › topics › ai-agents › article
[ARTICLE · art-152947] src=gist.github.com ↗ pub= topic=ai-agents verified=true sentiment=↑ positive

/better: an agent skill that asks for a deeper version of a task, not a bigger one. Drop into ~/.claude/skills/better/ or ~/.agents/skills/better/

A developer released /better, an agent skill that installs into ~/.claude/skills/better/ or ~/.agents/skills/better/ and pushes a task to a deeper version of itself rather than a larger one. The skill, triggered only when a user explicitly types /better or $better, first asks what a top professional in the field would do, what raw source material they would start from, which methods and tools they would use, and what independent evidence would confirm the result, then rewrites the prompt accordingly. It distinguishes between improving a not-yet-started prompt and deepening work already in progress, pausing for user approval when the revised brief changes the goal or requires decisions, money or access.

by read2 min views1 publishedOct 19, 2026
name better
description Push the current task, or a prompt the user gives, to a deeper version of itself. Use only when the user explicitly types /better or $better.

I want a deeper version of this task, not a bigger one. Before rewriting it: (1) say what a top professional in this field would do here that I'm not asking for; (2) name the rawest material they would start from instead of what has already been digested for me; (3) name the specific methods and tools they would use; (4) say what, other than you or me, would show the result is right. Then rewrite my prompt to ask for exactly that.

First decide which case you are in.

Case A: the user gave you a prompt to improve. The text with the command is a prompt or task they have not started yet. Give your answers to (1) to (4), then the rewritten prompt. Do not carry the prompt out.

Case B: work is already in progress. The command came alone, or with feedback on work you have already done in this conversation ("not good enough", "too shallow"). "This task" is that work, and the feedback is part of the task.

  1. Give your answers to (1) to (4) for that task, in a few lines each.
  2. Rewrite the task's brief: the original request plus the user's feedback, deepened by your answers.
  3. Decide whether to . If the new brief changes the goal, widens the scope, or needs a decision, money or access that only the user can give, show it and wait for their go-ahead. Otherwise carry on.
  4. Do the work again to the new brief. Start from the raw material in your answer to (2), use the methods and tools in your answer to (3), and before you report, run the check in your answer to (4) and say what it showed.
── more in #ai-agents 4 stories · sorted by recency
── more on @/better 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/better-an-agent-skil…] indexed:0 read:2min 2026-10-19 · —