BYOT — Bring Your Own Tests

Monitor your production business flows, not just your code.

BYOT is an open-source production monitoring platform that executes your existing test binaries — Playwright, k6, custom scripts, anything — on a schedule inside isolated Docker containers, and alerts you when something actually breaks.

---

The problem CI/CD doesn't solve

"I can just run my Playwright tests in CI on a schedule" — that's the first objection. GitHub Actions cron jobs exist. So why BYOT?

1. CI runs on code changes. Production breaks between deploys.

Monday  9am  — deploy ships ✅
Monday  3pm  — Stripe silently changes their API
Monday 11pm  — payments are broken ❌
Tuesday 9am  — customer complains

Your CI never ran. Your tests never ran. Nobody knew. BYOT runs continuously, regardless of deploys.

2. CI tests your code. BYOT tests your dependencies.

Your code didn't change. But:

• A third-party API changed their response format
• Your CDN went down
• The database ran out of connections
• An SSL certificate expired
• An external payment provider had an outage

None of these show up in CI. All of them show up in BYOT.

3. CI is for developers. BYOT is for the business.

CI/CD  →  "did my code break?"
BYOT   →  "is my checkout working right now?"

Product managers, founders, and support teams care about the second question. They can't read CI logs. BYOT gives them a green/red answer.

4. Flaky tests kill CI adoption.

Teams disable E2E tests in CI because flaky tests block deploys. Those tests rot. BYOT's consecutive-failure threshold means flaky tests don't cause noise — so teams actually keep them running.

5. One incident pays for itself.

One missed P0 = hours of engineer time + customer churn. If BYOT catches one checkout failure before a customer does, it paid for itself 10× over.

---

Who it's for

Not solo developers. Not teams with one service.

Target: Teams with 5+ engineers, multiple services, external dependencies, and an existing E2E test suite that's gathered dust because "it's too flaky for CI." That's a real market, and they get immediate value.

---

Why not Datadog or Checkly?

Feature	BYOT	Datadog	Checkly
Use existing tests	✓	✗	✗
Self-hostable	✓	✗	✗
Open source	✓	✗	✗
Flaky test detection	✓	paid tier	limited
Custom runtimes	✓	✗	limited
Free tier	unlimited	5 tests	3 tests

Datadog does this — at $500+/month with a proprietary DSL you have to learn. BYOT does it for $29 or free, with the tests you already wrote.

---

Quick Start

Prerequisites: Docker, Docker Compose, Go 1.21+

1. Start the infrastructure

docker compose up -d

This starts PostgreSQL (port 5432) and MinIO (port 9000 / console 9001).

2. Configure environment

cp .env.example .env
# Edit .env if needed — defaults work out of the box with docker compose

3. Run the server

go run .
# Starts on :3000 — migrations run automatically

---

API

Health check

GET /health
→ { "status": "ok" }

Upload a test binary

POST /uploadBinary
Content-Type: multipart/form-data

Field	Type	Required	Description
binary	file	yes	Compiled test binary
runtime	string	yes	go or node
severity	string	yes	P0 P1 P2 P3

curl -X POST http://localhost:3000/uploadBinary \
  -F "binary=@./e2e/checkout.test" \
  -F "runtime=node" \
  -F "severity=P0"

# → { "id": "a3f8b2c1-...", "message": "binary uploaded successfully" }

Run a test

POST /run/{id}
Content-Type: application/json

{ "filename": "a3f8b2c1-...", "runtime": "node", "timeout": 60 }

Output is stdout/stderr from the container, timestamped per line:

2026-05-03T10:30:00Z [stdout] ✓ Login successful
2026-05-03T10:30:04Z [stdout] ✓ Cart updated
2026-05-03T10:30:06Z [stderr] ✗ Pay button not found

---

Severity levels

Level	Meaning
P0	Wake me up at 3am — production is down
P1	Critical path broken, escalate fast
P2	Degraded, alert during business hours (default)
P3	Nice to know — weekly digest

---

Runtimes

Runtime	Docker image	Notes
go	alpine	Executes compiled binary directly
node	node:18	Runs with node interpreter

More runtimes (Python, Ruby, Deno) are planned — contributions welcome.

---

Architecture

Client
  │
  ▼
HTTP Server (:3000)
  │
  ├─ POST /uploadBinary ──► MinIO  ({uuid}/binary)
  │
  └─ POST /run/{id}
       │
       ├─ Download binary from MinIO
       ├─ docker run --rm -v /tmp/...:/app/binary <image>
       ├─ Stream stdout + stderr (RFC3339 timestamps)
       └─ Upload logs to MinIO  ({uuid}/logs/{timestamp}.txt)

PostgreSQL
  ├─ tests       (uuid, runtime, severity, binary_url, created_at)
  └─ tests_runs  (uuid, test_id, status, duration_ms, log_url, ...)

Each execution runs in a fresh, ephemeral Docker container — no shared state between runs.

---

Self-hosting

The full stack is a single docker compose up. For production:

• Point DB_URL at a managed Postgres (RDS, Supabase, Neon, etc.)
• Point MinIO config at S3 or any S3-compatible store
• Run the Go binary behind a reverse proxy (nginx, Caddy)

---

Development

docker compose up -d  # start infra
air                   # hot reload (go install github.com/air-verse/air@latest)
go test ./...

---

License

MIT — see LICENSE.