You want Copilot to work across several files, run commands, and iterate on failures. Which feature is designed for this?
Why A is the answer
Agent mode
Agent mode takes multi-step actions across files and tools. Inline suggestions only complete what you are typing.
In detail
- Agent mode works across files, runs commands and iterates on failures.
- Review the full set of changes when it finishes.
Use case
An illustrative scenario, not a real incident.
A developer asks to rename a module, update imports and fix failing tests. Agent mode edits the files, runs the tests and works through the failures.
See it in different ways
flowchart LR A["Code suggestions"] --> B["Focused edits"] B --> C["Interactive chat"] C --> D["Planning"] D --> E["Autonomous agent work"]
mindmap
root((Copilot))
Editor
Inline suggestions
Chat
Agent mode
GitHub
Cloud agent
Code review
Spaces
Terminal
Copilot CLI
Plan modeflowchart TD
Q{"What do you need?"}
Q -->|"Finish this line"| A0["Inline suggestions"]
Q -->|"Ask a question"| A1["Chat"]
Q -->|"Multi-file change with commands"| A2["Agent mode"]Key point: Agent mode is for multi-step work across files.
Why the other options are wrong
- B. Content exclusions
- Content exclusions restrict what Copilot can read.
- C. Inline suggestions
- Inline suggestions complete code where you type.
- D. Pull request summaries
- Summaries describe a change, and do not make one.