Skills vs slash commands: automatic or on request
A slash command runs when you type it; a skill is picked when a request matches. How to decide which fits a workflow, and when to have both.
The core difference is who decides. You decide when a slash command runs. Claude decides when a skill applies, based on its description.
| Slash command | Skill | |
|---|---|---|
| Runs when | You type it | A request matches its description |
| Best for | Rituals you start: release notes, a review pass | Standards that should always apply |
| Failure mode | Forgetting to run it | Not triggering, or triggering too often |
A good pairing: a slash command that starts a release, and a skill that holds the rules for release notes, which applies whether or not the command started it. If a skill must apply without fail, choose its description carefully: how to write a skill description.
Questions
What is a custom slash command?
A saved prompt you run on demand by typing its name with a slash. It does exactly what it says, every time you invoke it, and never on its own.
When should something be a skill instead?
When you want it applied automatically whenever the work calls for it, without remembering to invoke it.
Can a workflow use both?
Yes. A slash command can kick off a task on request, while skills supply the standards that apply during it.