open source · self-hostable · active-development

Your tests.
Running in production.

Upload your existing Playwright, k6, Node, Python or Go test. BYOT runs it on a schedule, isolated in Docker, and pages PagerDuty or Slack when something actually breaks.

Open Playground → View on GitHub
byot — live execution log
# Upload your existing test (or use the web UI)
$ curl -X POST /uploadBinary -H "Authorization: Bearer $TOKEN" \
-F binary=@checkout.js -F runtime=node -F name=checkout \
-F cron="*/30 * * * *" -F webhook_url=$SLACK_WEBHOOK
✓ Uploaded · scheduled every 30m · webhook configured
→ Test ID: a3f8b2c1 · node runtime

# 30 minutes later — runs automatically
2026-05-03T10:30:00Z [stdout] Starting checkout flow...
2026-05-03T10:30:02Z [stdout] ✓ Login successful
2026-05-03T10:30:04Z [stdout] ✓ Cart updated
2026-05-03T10:30:06Z [stderr] ✗ Pay button not found
2026-05-03T10:30:06Z [stderr] Test FAILED · 3rd consecutive failure
→ Failure webhook sent after 3 consecutive failures
0
lines to rewrite
<5m
to first alert
5
runtimes supported
100%
open source core
features
Built for engineers
who hate false alarms.

Alert only on failures that persist. Run your tests in a sandbox, with secrets handled properly. No DSL to learn. No lock-in.

◈
Bring your own test
Upload a Playwright project, a k6 script, or a Node, Python or Go program. If it runs in one of those, BYOT runs it.
◷
Smart scheduling
Every 5 minutes or once a day. Cron syntax. Tests run isolated in Docker containers every time.
◉
Flaky test detection
Fail 3 times in a row before alerting. Configurable thresholds per test. No 3am noise.
≡
Streaming logs
Every stdout and stderr line captured with timestamps. Full history stored in S3/MinIO.
◈
Failure thresholds
Alert only after N consecutive failures, and get a recovery ping when it passes again. No flaky-test noise.
⌘
PagerDuty and Slack alerts
Add a PagerDuty integration key and BYOT opens one incident per test, then resolves it when the test passes again. Or set a Slack-compatible webhook URL for a failure message and a recovery message. Use either or both.
⚿
Secrets, encrypted
Give each test its own environment variables. Mark API keys as secret: they're encrypted at rest, never shown again, and scrubbed from your logs.
▣
Sandboxed by default
Every run is a throwaway container: non-root, no extra privileges, read-only filesystem, CPU and memory limits. Network is off unless the runtime needs it.
⇉
Runs in parallel
Ten monitors due at the same minute don't queue behind each other. Concurrency is configurable to fit your host.
◫
Screenshots on every run
Playwright screenshots are collected and stored with the run, so you can see what the page looked like when it failed.
↻
Self-cleaning
Hung tests are killed at their timeout, runs abandoned by a restart are failed automatically, and old runs, logs and screenshots are pruned on a schedule.
✓
Open to read, sign-in to change
Anyone can look at the dashboard. Uploading, running, scheduling, deleting and editing secrets all require a sign-in.

how it works
Up in 4 steps.
01
Upload
Push your compiled test binary. Any language, any framework.
02
Configure
Set schedule, timeout, environment variables and secrets, and PagerDuty or Slack. Takes 2 minutes.
03
Run
BYOT executes in isolated Docker containers and streams every log line.
04
Alert
Only real failures wake you up. Smart thresholds filter the noise.

get started
Running in 5 minutes.

All you need is Docker. The database, storage and migrations set themselves up on first start.

try it locally
# 1. Download the compose file
$ curl -O https://raw.githubusercontent.com/alexanderritik/BYOT/main/docker-compose.yml
# 2. Start BYOT, Postgres and MinIO
$ docker compose up -d
✓ Open http://localhost:3000/app and upload a test
Going to production?
  1. Pin a version and generate passwords and a secrets key in a .env file.
  2. Run docker compose up -d on a server dedicated to BYOT.
  3. Keep it on a private network, or turn on sign-in and put HTTPS in front.
Full production guide →

comparison
Why not Datadog or Checkly?

They're mature hosted products, and Checkly runs Playwright tests well. BYOT is the same idea, fully open source, on your own servers, across more test runtimes.

Feature BYOT Datadog Checkly
Open source (MIT) ✓ ✗ CLI only
Runs entirely on your infrastructure ✓ private locations, hosted control plane private locations, hosted control plane
Runtimes Playwright, k6, Locust, Node, Python, Go binaries Datadog browser & API tests Playwright, API checks
Pricing free your hardware per test run per check run
Managed service, global locations, SLAs ✗ ✓ ✓

vs ci/cd
"Can't I just use
GitHub Actions cron?"

Yes. But CI runs on code changes. Production breaks between deploys.

◷
The 3pm problem
Deploy ships at 9am ✓. Stripe silently changes their API at 3pm. Payments break at 11pm. Customer complains Tuesday morning. Your CI never ran — it had no code change to trigger on.
◈
CI tests your code. BYOT tests your dependencies.
Third-party API change. CDN outage. Database out of connections. Expired SSL cert. None of these show up in CI. All of them show up in BYOT.
◉
CI is for developers. BYOT is for the business.
CI answers "did my code break?" BYOT answers "is my checkout working right now?" — the question your product manager and founder actually care about.

Stop finding out
from your customers.

Deploy BYOT in 5 minutes. Your existing tests start monitoring production today.

Open Playground→ ★ Star on GitHub