Dependency Updates That Arrive Tested
On a schedule you set, an agent updates your libraries, runs the tests, and opens a pull request, so updates arrive ready to review.
Updates Pile Up Until They're a Project
Small version bumps are easy to put off, and they add up. Automated update tools open the pull request, but when an update breaks a test or needs a code change, the work lands back on an engineer. Skipped updates turn into a large, risky upgrade later.
An Agent Does the Update and the Follow-Up
An automation runs on the schedule you choose. The agent updates dependencies in an isolated environment, runs the tests, and adjusts code when an update needs it. It opens a pull request with the changes and the test results, and a person reviews it like any other PR.
- "@tanstack/react-query": "5.59.0",
+ "@tanstack/react-query": "5.62.0",
- "zod": "3.23.8",
+ "zod": "3.24.1",
Tests passed, PR opened
Common Questions
How often does it run?
On any schedule you set with a cron expression, or once at a chosen date and time.
What happens when an update breaks the tests?
The agent works on the failure like any other task, and adjusts the code when the fix is clear. If it can't, it reports why in the pull request or session.
Does it merge updates automatically?
No. Every update arrives as a pull request for a person to review.
Which models can it use?
Any major provider, plus local and private models. Bring your own keys or subscription, and there's no markup on tokens.
Related: Automations docs, Security patches.
Keep Your Attention for the Hard Problems
Start with one repo and one recurring task.