A Claude Code plugin with eight very small skills. Each one is a phrase I type often. There is no logic in any of them. The whole skill is the sentence.
The point is that a short prompt you repeat every day should be one word instead.
| Command | What it says |
|---|---|
/nice:nice |
Nice, anything you would've benefited from in terms of knowledge/tooling to make this easier? |
/nice:what-do |
What should we do? |
/nice:do-it |
/plan ok let's do it |
/nice:what-info |
Nice, what other information would be useful for you to be very ready to handle any deep issue that arizes, and to figure it out very fast, and potentially solve it fast as well, if it's an extreme emergency? |
/nice:toc |
show me a brief table of contents of what you did in the last following number of hours: |
/nice:fastest |
What is the aggressively fastest path to it |
/nice:smallest |
Make the smallest complete change and avoid adjacent cleanup or improvements. |
/nice:pls |
Asks for an answer that is detailed but short and easy to read. |
/nice:toc takes a number, so you type /nice:toc 3.
Clone it into your Claude Code skills folder.
git clone https://github.com/testy-cool/nice.git ~/.claude/skills/niceThen restart Claude Code, or run /reload-plugins.
Codex, Gemini and Pi read plain skill folders and do not know about plugins. Link the individual skills into their skill folders instead.
for a in codex gemini pi; do
for s in nice what-do do-it what-info toc fastest smallest pls; do
ln -s ~/.claude/skills/nice/skills/$s ~/.$a/skills/$s
done
doneThere the names have no prefix, so you type /what-do rather than /nice:what-do.
Copy any folder under skills/, change the folder name, and write your sentence
in the body. The folder name is the command name. The description in the
frontmatter is what the model reads to decide the skill applies, so say when it
should fire. Nothing else is needed.
MIT.
