Story · Test automation

End-to-end tests that report live in Slack

An end-to-end test system where every run narrates itself in Slack: what is queued, what passed, how long it took — and where the evidence is.

For a streaming startup I built the end-to-end testing system for their web products — and gave it a strong Slack integration, so the whole team could watch tests perform live instead of digging through CI logs.

Every pull request got its own test run against its own preview deployment. A bot posted the run to Slack and kept updating the same message: queued tests, live progress, then a green tick and a timing for each test as it finished — with links to the PR, the ticket, and the exact commit.

Slack message from a test automation bot showing a completed end-to-end test run with green check marks and per-test timings
A finished run, as the team saw it: every test green, timed, and traceable to a commit. (“Which means that the features tested during this test run are not obviously broken.”)
A run completing live: each test reports its result and timing into the same Slack message as it finishes.

When something failed, the evidence was already collected: HTML logs, HAR files, traces, and a screen recording of the failing run, uploaded and linked from the same thread. Debugging started from the artifacts, not from trying to reproduce.

The effect was social as much as technical: test results stopped being something you went looking for and became something the team talked about — in the channel, in the open, in real time.

← All stories

Have a system that needs this kind of attention?

Tell us where your current testing approach breaks down. We will help you define a practical path forward.

Talk to WhileOne