GitHub Copilot has evolved beyond inline completions into a configurable AI development platform. Master custom instructions, Copilot Extensions, agent mode, Knowledge Bases, and workspace context for team-scale AI development workflows.
0 / 5 completed
1 / 5
A developer wants to add company-specific context to GitHub Copilot suggestions across their repository. Which Copilot feature allows defining project-level AI instructions?
The .github/copilot-instructions.md file lets teams define project-level custom instructions that Copilot Chat appends to every conversation in that repository. This enables contextual guidance — preferred libraries, coding standards, architecture decisions — without each developer needing to paste context manually.
2 / 5
A team builds a GitHub Copilot Extension that integrates with their internal API. What type of extension connects Copilot Chat to external services via a webhook?
A Copilot Agent Extension (formerly Copilot Extension of type "agent") connects Copilot Chat to an external service via a webhook endpoint. When a user invokes the extension in chat (e.g., @my-extension what is the status?), GitHub sends a Server-Sent Events request to the extension's webhook, which responds with streaming content.
3 / 5
In GitHub Copilot's agent mode, how does Copilot differ from standard inline completions?
Agent mode elevates Copilot from a completion tool to an autonomous coding assistant. It can plan multi-step tasks, read and edit multiple files, run terminal commands, interpret error output, and iterate — all within the IDE. Unlike inline completions, agent mode operates on the broader codebase context to complete higher-level tasks.
4 / 5
A Copilot Knowledge Base is configured for an organization. What does it provide?
A Knowledge Base allows Copilot Enterprise customers to index internal documentation (Markdown files, wikis, etc.) stored in GitHub repositories. Copilot Chat can then retrieve relevant sections and cite them in answers, grounding responses in company-specific knowledge rather than relying solely on general training data.
5 / 5
A Copilot Chat workspace context (@workspace) is used in a prompt. What does this context include?
The @workspace context scopes Copilot Chat's understanding to the entire open repository. Copilot indexes the workspace and can answer questions like "where is authentication handled?" or "generate a test for this function based on similar tests in the project" by retrieving relevant code from across all files.