Consider merging Skills and Slash Commands

[ykdojo]

Skills and Slash Commands seem to function almost identically:

• Both can be invoked by the user with a slash prefix
• Both can be invoked by Claude itself
• Both load instructions into the conversation

The main difference appears to be intent (skills = Claude-invoked primarily, commands = user-invoked primarily), but in practice they work the same way.

Would it simplify things to merge these into a single feature? Perhaps with a flag to control whether it appears in autocomplete or is primarily model-invoked?

[github-actions[bot]]

Found 3 possible duplicate issues:

1. https://github.com/anthropics/claude-code/issues/11459
2. https://github.com/anthropics/claude-code/issues/5624
3. https://github.com/anthropics/claude-code/issues/10246

This issue will be automatically closed as a duplicate in 3 days.

• If your issue is a duplicate, please close it and 👍 the existing issue instead
• To prevent auto-closure, add a comment or 👎 this comment

🤖 Generated with Claude Code

[dicksontsai]

The main difference appears to be intent (skills = Claude-invoked primarily, commands = user-invoked primarily)

Yes. We think this difference is significant enough to keep the two concepts separate, especially since it would help with how Claude authors skills and slash commands.

in practice they work the same way

We are definitely looking to eliminate this, but it's tricky since users may depend on the fact that they work the same way.

Both load instructions into the conversation

The way they load instructions may diverge over time.

@dhollman is very knowledgeable in and invested in this topic!

[tonydehnke]

We are definitely looking to eliminate this, but it's tricky since users may depend on the fact that they work the same way.

Both load instructions into the conversation

The way they load instructions may diverge over time.

@dhollman is very knowledgeable in and invested in this topic!

I'd like a way to call a skill in a prompt either via something like /skill:some-skill-name or @skill:some-skill-name etc