← Back to Discover

Cursor

Coding agent that takes on development tasks, runs cloud agents in parallel, and works from the terminal, Slack and GitHub with models from several providers.

Coding agents
Visit website ↗

What it does

Cursor describes itself as a coding agent for building software. You hand it a task, it explores the codebase and carries the work through, and you stay in charge of decisions and review. Cloud agents run on their own computers to build, test and demo features end to end, and many can run in parallel on long tasks. Automations let you set up always-on agents that run on schedules or triggers to build, maintain and fix software. It is meant to show up where developers already work: in the terminal, which has a curl-based installer, in Slack for team collaboration, and in GitHub for pull request review. A model picker lets you choose among models from OpenAI, Anthropic, Gemini, SpaceXAI and Cursor, or leave the choice on Auto.

Capabilities

  • Agents that carry a task from request to tested, demoed work
  • Cloud agents that run in parallel
  • Automations: always-on agents on schedules or triggers
  • Use from the terminal, Slack and GitHub pull request review
  • Model choice across OpenAI, Anthropic, Gemini, SpaceXAI and Cursor

When to use it

  • Handing a well-defined feature or fix to an agent and reviewing the result
  • Running several coding tasks at once in cloud agents
  • Setting up recurring maintenance work that runs on a schedule or trigger
  • Working with an AI agent from the terminal, Slack or GitHub