Currents

Currents gives AI agents the test results your CI records, from Playwright, Vitest, Jest and more.

Self-healing tests

• Ask why a run failed and get the errors, stack traces, screenshots and traces behind it
• Step through a run's Playwright trace, action by action, to see what the page did
• Find the root cause and fix the test, flaky ones included
• Compare a pull request's runs with main to see what the change broke

Proof of work for any change

• Before-and-after screenshots, videos, Playwright traces and DOM snapshots from CI
• A demo the agent records in a browser, saved as a run
• Evidence links posted in the pull request, so reviewers see the change working

Test suite health

• Rank your flakiest and slowest tests and spec files
• See which errors break the most tests
• Quarantine, skip or tag tests with rules, and review what they affected

Day-to-day CI work

• Reproduce a bug from a failed run and verify the fix
• Open an issue for a failure in your issue tracker
• Cancel, reset or re-check runs without leaving the chat

For teams whose CI slows them down: flaky tests that block merges, failures nobody has time to dig into, and AI-written pull requests that are hard to trust without seeing the change work. Works with GitHub, GitLab, Slack, Jira, Linear and more. You need a Currents account.