Dagger Shell with An (actually) Autonomous QA Agent

Lev’s back with a new and improved QA agent demo—this time, it’s (actually) autonomous. After a previous single-run setup that required manual kicks, this agent now runs entirely on its own via hourly cron jobs. It uses Dagger’s new Environments API and LLMs to install, test, analyze, and publish QA reports. Watch to see how far full autonomy can go and where it still needs a nudge. Repo with all the code: https://github.com/levlaz/agent-playground Sample Dagger Cloud Trace: https://v3.dagger.cloud/levlaz/traces/8500d18e40016fe8d99d0ad62719041b