Configuration

Every env var the server reads. None are required, so the defaults will get you a running instance. Set what you need; ignore what you don't.

General

Name	What it controls	Default
PORT	HTTP port	3001
HOST	Bind address	0.0.0.0
API_TOKEN	Turns auth on (and authorizes first-run setup). See Authorization and the Authentication page.	unset (open)
AUTH_SECRET	Encrypts stored secrets (e.g. GitHub sync tokens).	dev fallback
UI_AUTH_EXPIRE_HOURS	Idle session lifetime, in hours (sliding; capped at 30 days absolute)	12
COOKIE_SECURE	Secure flag on auth cookies. Set false only for plain-HTTP/Localhost.	true
ROOT_USERNAME	Break-glass admin username (recovery only; needs ROOT_PASSWORD)	unset
ROOT_PASSWORD	Break-glass admin password (recovery only; needs ROOT_USERNAME)	unset
DATA_STORAGE	fs / s3 / azure. See Storage.	fs
GITHUB_TOKEN	Fallback token for GitHub Sync configs without one of their own	unset

Authorization

Set API_TOKEN to any non-empty value and auth turns on: the UI requires login, the API requires a session cookie or an API key. Leave it unset and the server runs in open mode - anyone with the URL gets full access, which may or may not be what you want. API_TOKEN is also the secret that authorizes first-run setup of the initial admin. More details - Authentication page.

Set AUTH_SECRET to a long random value so encrypted secrets (e.g. GitHub sync tokens) survive restarts. Without it the server falls back to a fixed dev key, which is unsafe for production. Rotating it invalidates anything it encrypted.

# generate one
openssl rand -base64 32

S3

Set DATA_STORAGE=s3 plus:

Name	What it controls	Default
S3_ENDPOINT	Hostname only, no scheme
S3_REGION	Region (auto for Google Cloud)
S3_ACCESS_KEY	Access key
S3_SECRET_KEY	Secret key
S3_BUCKET	Bucket name	playwright-reports-server
S3_PORT	Custom port (self-hosted)
S3_USE_SSL	Use HTTPS to the S3 endpoint (false for local MinIO over plain HTTP)	true
S3_BATCH_SIZE	Concurrent requests during cleanup	10
S3_MULTIPART_CHUNK_SIZE_MB	Multipart upload chunk size, MB	25

When you pick S3, Litestream replicates the SQLite metadata DB to the same bucket automatically. See Storage -> S3 for the GCS quirks if you went the Google route.

Azure

Set DATA_STORAGE=azure plus:

Name	What it controls	Default
AZURE_ACCOUNT_NAME	Storage account
AZURE_ACCOUNT_KEY	Account key
AZURE_CONTAINER	Container name (created if missing)	playwright-reports-server
AZURE_BATCH_SIZE	Concurrent requests during cleanup	10

Litestream replicates the SQLite metadata DB to the same container automatically. See Storage -> Litestream.

Checking what's actually set

The Settings page shows resolved environment. Useful after editing .env and forgetting to restart, which everyone has done at least once.

See also

• Storage
• GitHub Sync
• LLM analysis
• Deployment