[BUG] Selecting a skill from autocomplete always resolves to first plugin when multiple plugins share the same skill name
Preflight Checklist
- [x] I have searched existing issues and this hasn't been reported yet
- [x] This is a single bug report (please file separate reports for different bugs)
- [x] I am using the latest version of Claude Code
What's Wrong?
When two installed plugins register a skill with the same short name (e.g., both plugin-a and plugin-b have a jira skill), the autocomplete doesn't correctly show both entries with their plugin prefix (only the description has the plugin info):
/jira (plugin-a) Description A...
/jira (plugin-b) Description B...
However, selecting the second entry (plugin-b) still resolves and executes the first plugin's skill (plugin-a:jira). The only workaround is manually typing the fully qualified name /plugin-b:jira, which has no autocompletion.
What Should Happen?
Selecting a specific entry from the / autocomplete list should invoke that plugin's skill using its fully qualified name (e.g., /plugin-b:jira).
Error Messages/Logs
No error — the wrong skill silently executes.
Steps to Reproduce
- Install two plugins that each expose a skill with the same name (e.g.,
jira) - Type
/jirain the input — both options appear in autocomplete - Select the second entry from the list
- Observe that the first plugin's skill is always executed
- Manually typing
/plugin-b:jiracorrectly invokes the intended skill, confirming the resolution logic works — the bug is in the autocomplete picker not passing the qualified name
Claude Model
Not sure / Multiple models
Is this a regression?
Yes, this worked in a previous version
Last Working Version
_No response_
Claude Code Version
2.1.37
Platform
Anthropic API
Operating System
macOS
Terminal/Shell
Terminal.app (macOS)
Additional Information
The fully qualified name (/plugin-b:jira) works correctly when typed manually but has no autocompletion support. The autocomplete picker appears to discard the plugin namespace when dispatching the selected skill.
Additionally, before skills and commands were merged, skills were autodiscovered and invoked automatically. The logs would show the full skill path, making it clear which skill was being used. After the merge (skills became slash-invocable like commands), the namespace seems to be discarded during resolution, and there is no longer any indication in the logs that a skill is being used or which one was selected.
This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗