Glossary
What is agent mode?
Agent mode is a setting in an AI product that lets the model act on its own across several steps, using tools to finish a task instead of only replying. Coding tools such as Cursor and GitHub Copilot, and assistants such as ChatGPT, offer agent modes. Products can also run inside your agent.
Chat vs agent mode
| Chat | Agent mode | |
|---|---|---|
| Output | An answer | Completed task |
| Tools | Rarely | Files, terminal, browser, APIs |
| Steps | One | Many, chosen by the model |
| Oversight | Read the reply | Approve actions |
Products built for agent mode
Some products skip their own interface and run inside the user's existing agent: a repository or package with instructions, data and scripts that Claude Code, Codex or Cursor can follow. The user keeps their own model, browser and accounts. AgentLaunch works this way: the launch kit holds the directory catalog and recorded submit paths, and your agent does the submitting.
Keeping it safe
Agent modes should ask before running commands, spending money, entering credentials or doing anything irreversible, and should log their actions. See human in the loop.
Example
In agent mode, a user asks for a product to be submitted to developer directories. The agent finds the right ones in the kit's catalog, opens each submit page in the user's browser, fills the forms and asks before a paid option.
FAQ
Is agent mode the same as an AI agent?
Agent mode is a product setting that turns an assistant into an agent for that task.
Is agent mode safe?
With approvals for sensitive actions and logs, it's manageable; review what it does.
Which tools have agent mode?
Cursor, GitHub Copilot, ChatGPT and others offer agent modes.