Cursor Launches Agent SDK for Building and Running AI Bots

At a glance

Cursor has officially released its Agent SDK. Developers can build agents as code folders, test them with evaluation tools, and deploy them to Cursor or their own servers. The tool is designed to make it easier for developers to build and run AI bots themselves.

Key points

1. 1

   Agents are built as a single project folder that includes instructions, tools, and technical procedures.

2. 2

   Bots can wake up and work based on events from GitHub, GitLab, Bitbucket, and Slack, or on a set schedule.

3. 3

   Evaluations, which check a bot's behavior with preset messages, can be run locally or in CI (automated build and test environments).

4. 4

   With a single command, developers can resend a real pull request as a fake webhook signal to test the bot.

Why it matters

By combining the process of building and verifying AI bots into one tool, it gives people who want to run automation agents in their development work a practical option.