Microsoft GH-300 - GitHub Copilot Exam

Restricts Copilot to use only external databases for generating code suggestions.

Allows Copilot to generate suggestions based only on the prefix of the code.

Ignores both the prefix and suffix of the code, focusing only on user comments for context.

Improves suggestions by considering both the prefix and suffix of the code, filling in the middle part more accurately.

Inclusiveness

Fairness

Reliability and Safety

Privacy and Security

It primarily relies on the model's training data to generate responses.

It primarily uses search results from Bing to generate responses.

It combines its training data set, code in user repositories, and external sources like Bing to generate responses.

It uses user-provided documentation exclusively to generate responses.

By opening the relevant tabs in your IDE

By adding relevant code snippets to your prompt

By adding the important files to your .gitconfig

By adding the full file paths to your prompt of important files

Suggestions matching public code to 'blocked'

Enforce blocking of MIT or GPL licensed code

You need to check code suggestions yourself before accepting

Enable GitHub Copilot license checking

Utilizing GitHub Copilot's ability to create fictitious information from patterns in training data.

Leveraging GitHub Copilot's ability to independently initiate and manage data storage services.

Utilize GitHub Copilot's capability to directly access and use databases to create sample data.

Leveraging GitHub Copilot's suggestions to create data based on API documentation in the repository.