GitHub's Copilot app organizes AI agentAn AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.Full definition → work into sessions that combine the task, code changes, and review tools. A session can start from an issue, a prompt, or an existing pull request, with its own workspace for files, branches, and conversation. Developers can inspect changes, preview the application in a browser, and run terminal checks before merging. The useful capability is keeping delegated work and the evidence needed to review it together. These capabilities are described on the product page.
Terms in this piece · Glossary
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
MCP — The Model Context Protocol — an open standard that lets any AI assistant plug into any tool or data source without custom integration code.
agent skill — A reusable instruction file that teaches an agent how to do one job well — the procedure, the tools, and what counts as done.
