Skip to content

Commit d29f11b

Browse files
hubwriterCopilot
andauthored
Add note about custom instructions varying behavior (#58511)
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
1 parent 4a3e808 commit d29f11b

File tree

2 files changed

+14
-0
lines changed

2 files changed

+14
-0
lines changed

content/copilot/concepts/prompting/response-customization.md

Lines changed: 12 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -26,6 +26,8 @@ category:
2626

2727
{% data variables.product.prodname_copilot %} can provide responses that are tailored to your personal preferences, the way your team works, the tools you use, or the specifics of your project, if you provide it with enough context to do so. Instead of repeatedly adding this contextual detail to your prompts, you can create custom instructions that automatically add this information for you. The additional information is not displayed, but is available to {% data variables.product.prodname_copilot_short %} to allow it to generate higher quality responses.
2828

29+
{% data reusables.copilot.custom-insts-nondeterministic %}
30+
2931
## Types of custom instructions
3032

3133
There are three main types of custom instructions that you can use to customize {% data variables.product.prodname_copilot_short %} responses on the {% data variables.product.github %} website:
@@ -122,6 +124,8 @@ There are two types of files you can use to provide context and instructions to
122124

123125
While custom instructions help to add codebase-wide context to each AI workflow, prompt files let you add instructions to a specific chat interaction.
124126

127+
{% data reusables.copilot.custom-insts-nondeterministic %}
128+
125129
## About repository custom instructions
126130

127131
You can use three types of repository custom instructions in {% data variables.product.prodname_vscode_shortname %}:
@@ -199,6 +203,8 @@ The following examples demonstrate how to use prompt files.
199203

200204
{% data variables.product.prodname_copilot %} can provide responses that are tailored to the way your team works, the tools you use, or the specifics of your project, if you provide it with enough context to do so. Instead of repeatedly adding this contextual detail to your prompts, you can create a custom instructions file in your repository that automatically adds this information for you. The additional information is not displayed in the chat input box, but is available to {% data variables.product.prodname_copilot_short %} to allow it to generate higher quality responses.
201205

206+
{% data reusables.copilot.custom-insts-nondeterministic %}
207+
202208
## About repository custom instructions
203209

204210
In {% data variables.product.prodname_vs %}, repository custom instructions consist of a single file, `.github/copilot-instructions.md`, that you create in a repository. The instructions you add to the file should be short, self-contained statements that add context or relevant information to supplement a {% data variables.product.prodname_copilot_short %} prompt.
@@ -246,6 +252,8 @@ There are two types of files you can use to provide context and instructions to
246252

247253
While custom instructions help to add codebase-wide context to each AI workflow, prompt files let you add instructions to a specific chat interaction.
248254

255+
{% data reusables.copilot.custom-insts-nondeterministic %}
256+
249257
## About repository custom instructions
250258

251259
In JetBrains IDEs, repository custom instructions consist of a single file, `.github/copilot-instructions.md`, that you create in a repository. The instructions you add to the file should be short, self-contained statements that add context or relevant information to supplement a {% data variables.product.prodname_copilot_short %} prompt.
@@ -337,6 +345,8 @@ For a curated collection of examples, see [AUTOTITLE](/copilot/tutorials/customi
337345

338346
{% data variables.product.prodname_copilot %} can provide responses that are tailored to the way your team works, the tools you use, or the specifics of your project, if you provide it with enough context to do so. Instead of repeatedly adding this contextual detail to your prompts, you can create a custom instructions file in your repository that automatically adds this information for you. The additional information is not displayed in the chat input box, but is available to {% data variables.product.prodname_copilot_short %} to allow it to generate higher quality responses.
339347

348+
{% data reusables.copilot.custom-insts-nondeterministic %}
349+
340350
## About repository custom instructions
341351

342352
In Xcode, repository custom instructions consist of a single file, `.github/copilot-instructions.md`, that you create in a repository. The instructions you add to the file should be short, self-contained statements that add context or relevant information to supplement a {% data variables.product.prodname_copilot_short %} prompt.
@@ -382,6 +392,8 @@ There are two types of repository custom instructions you can use to provide con
382392
* **Workspace custom instructions** which apply to all projects in a workspace, and allow you to specify workspace-wide instructions and preferences, in a single file.
383393
* **Project custom instructions** which apply to a specific project.
384394

395+
{% data reusables.copilot.custom-insts-nondeterministic %}
396+
385397
## About repository custom instructions
386398

387399
**Workspace custom instructions** apply to all projects in a workspace, and allow you to specify workspace-wide instructions and preferences. You can use workspace custom instructions to provide context and instructions to {% data variables.product.prodname_copilot_short %} in Eclipse.
Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,2 @@
1+
> [!NOTE]
2+
> Due to the non-deterministic nature of AI, {% data variables.product.prodname_copilot_short %} may not always follow your custom instructions in exactly the same way every time they are used.

0 commit comments

Comments
 (0)