Skip to content

Commit eb7e358

Browse files
authored
Merge pull request #119 from github/update-skills-multi-invoke-v1.0.44-d4c13763c489a36b
[bot] docs(ch05): document multi-skill invocation and mid-input slash commands (v1.0.44)
2 parents 6b69e2e + 5ed858b commit eb7e358

1 file changed

Lines changed: 14 additions & 0 deletions

File tree

05-skills/README.md

Lines changed: 14 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -101,6 +101,20 @@ While auto-triggering is the primary way skills work, you can also **invoke skil
101101

102102
This gives you explicit control when you want to ensure a specific skill is used.
103103

104+
#### Combining Multiple Skills in One Message
105+
106+
You can invoke **more than one skill in a single message**, and the skill slash command can appear anywhere in your prompt — not just at the beginning. This is handy when you want two different checks done in one go:
107+
108+
```bash
109+
> Check @samples/book-app-project/book_app.py with /code-checklist and also run /generate-tests for it
110+
111+
> Review the auth module /security-audit then /code-checklist the result
112+
```
113+
114+
Copilot will apply each named skill in the same response, saving you from sending multiple separate messages.
115+
116+
> 💡 **Tip**: Put the skill slash commands wherever they feel most natural in your sentence. You can put them at the start, middle, or end of your message.
117+
104118
> 📝 **Skills vs Agents Invocation**: Don't confuse skill invocation with agent invocation:
105119
> - **Skills**: `/skill-name <prompt>`, e.g., `/code-checklist Check this file`
106120
> - **Agents**: `/agent` (select from list) or `copilot --agent <name>` (command line)

0 commit comments

Comments
 (0)