The Agent SDK
Web page · https://cursor.com/blog/agent-sdk
First seen by Oriel on .
AI summary
Cursor’s generally available Agent SDK lets software teams define, test, and deploy event-driven agents as code, either on Cursor infrastructure or as containerized Node processes on their own infrastructure.
- Agents combine Markdown prompts, TypeScript tools, external service connections, and evals in version-controlled projects that teams can inspect and review.
- They can react to pull request events from GitHub, GitLab, or Bitbucket, assess risk, route reviews, enforce policies, and approve compliant changes.
- Cursor’s team built kb-reviewer in about a day; it checks every main-codebase PR against recorded lessons, flags contradictions, and cites supporting notes.
- Built-in evals and event replay support iteration, while specialized agents can delegate tasks or use parallel subagents for more complex workflows.
- agent-sdk deploy builds and hosts an agent with a URL; local development includes a playground, logs, evals, and skills for coding agents.
#software agents #developer tools #typescript #code review
Generated by gpt-5.6-sol. Check the original for details.
Metric trends
Last 30 days · Up to 120 recorded samples per source and metric, with reposts kept separate · First and latest retained
Points
Score reported by the source.
- Hacker News: 139, latest observation ; +130 over observed period since
Comments
Comment count reported by the source.
- Hacker News: 130, latest observation ; +129 over observed period since
Aliases
| Source | External ID | URL |
|---|---|---|
| Hacker News | 50273731 | https://cursor.com/blog/agent-sdk Read discussion |