Sonarly

Say a Sentry error fires at 2 a.m. and three Datadog alerts follow for the same fault. Sonarly takes those alerts, groups the duplicates, and works out the root cause so you do not have to dig through logs yourself.

Sonarly is an AI agent for monitoring and fixing production bugs that triages alerts and writes the fix with full system context.

It connects to observability tools such as Sentry and Datadog and pulls in errors along with logs, traces and metrics.

From there it investigates across the codebase and services, then opens a pull request on GitHub that accounts for how the code runs in production.

Its analysis is posted to Slack or Discord, so the team reads the findings where they already talk and does not need to switch tools.

The work is built on Claude Code and Opus 4.6.

It makes sense for engineering teams that are tired of noisy alert queues and want a drafted fix to review rather than a blank ticket.