Guiding Copilot with context

Software Development with GitHub Copilot

Thalia Barrera

AI Engineering Curriculum Manager, DataCamp

The importance of context

The #codebase variable

#codebase: Performs semantic code search across an entire project

Example prompt:

How is user authentication handled? #codebase

Result:

🔍 Scans entire codebase
💻 Includes relevant logic, function calls, or config settings in its answer

How Copilot understands your codebase

Index types:

Local index: Stored on your machine
Remote index: For repositories hosted on GitHub

Checking your index type

The #changes variable

#changes: References most recent source control changes

Example prompt:

Will any of these changes break the login flow? #changes

Use when:

💾 Evaluating code modifications

The #selection variable

#selection: References highlighted blocks of code

Example prompt:

Can you refactor this? #selection

Use when:

🔧 Making quick edits
📝 Creating targeted prompts

The #fetch variable

Documentation 📚, API reference 💻, blog post 📰

#fetch: Pulls the contents of a web page

Example prompt:

Summarize this documentation for me #fetch example.com/docs

Result:

🌐 Fetches a web page and uses it for context

The #files variable

#files: Attaches specific files to a prompt

Example prompt:

Can you check this file for error handling issues? #files:utils/data_loader.py

Result:

🔍 Analyzes the contents of the referenced file
📄 Includes it as part of the prompt context

Other chat variables and tools

Variable/tool	Description
`#search`	Enable searching for files in the current workspace.
`#testFailure`	Get unit test failure information. Useful when running and diagnosing tests.
`#problems`	Add workspace issues and problems from the Problems panel as context. Useful while fixing code or debugging.
`#terminalLastCommand`	Get the last run terminal command and its output.
`#githubRepo`	Perform a code search in a GitHub repo.

¹ https://code.visualstudio.com/docs/copilot/reference/copilot-vscode-features#_chat-tools

Building a progressive chat interaction

Single, complex prompt:

Add PostgreSQL integration with authentication, persistence, sessions, error handling, and connection pooling

Progressive complexity flow:

What's the best approach for adding database persistence? #codebase
Implement the database connection using the suggested approach
Help me with the schema design for the Task and the User classes
Now add session management and basic error handling
...

Starting a new chat to clear context

Let's practice!

Software Development with GitHub Copilot