Adds the bin/worker-notifications entrypoint, Dockerfile chmod, and
docker-compose service definitions for the new Notifications worker
without disturbing the existing Mails worker, and re-adds the Mails
queue/class constants on Event so the legacy callers still resolve.
The full mail->notification swap (caller migration, Mails removal,
worker rename) will land as a follow-up.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Drop the global _APP_NOTIFICATIONS_WEBHOOK_SECRET env var. There's no
analogous global webhook secret in Appwrite; the existing Webhooks
worker carries a per-webhook signatureKey on the webhook document.
Move the same pattern into the Notification event: each webhook
recipient may carry an optional signatureKey, which the worker
forwards to the Webhook adapter for HMAC-SHA256 signing. Recipients
without a key are delivered unsigned and a tag is logged for audit.
Registers the webhook signing secret in app/config/variables.php under a
new Notifications category, threads it through the worker container in
docker-compose.yml and tests/resources/docker/docker-compose.yml, and
adds an empty default to .env so operators see the knob alongside the
existing SMTP block.
When set, outbound webhook deliveries from the notifications worker
include an X-Appwrite-Webhook-Signature header carrying
sha256=<hex(hmac_sha256(timestamp.body, secret))>. When unset, the
worker delivers payloads unsigned — receivers must decide whether to
accept them.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
MongoDB's official Docker entrypoint uses a two-phase startup: a
temporary mongod for user/db init, then the real mongod. Under CI
resource pressure the port may not be released between the two
phases, causing mongod to exit with code 48 (address already in use).
Adding restart: on-failure:3 lets Docker handle the transient failure
natively. On restart the data directory already exists so the
entrypoint skips the two-phase init entirely, avoiding the race.
Add proper healthchecks to infrastructure services and use
condition: service_healthy for redis in appwrite's depends_on
so it waits for Redis to be ready before starting.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace the manual shell polling loop (until doctor > /dev/null) with a
proper Docker healthcheck on the appwrite service and `docker compose up
--wait`, which blocks until healthchecks pass. Also reverts the timeout
back to 3 minutes now that image pulls are cached.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Move adminer, redis-insight, mongo-express, and graphql-explorer to an
override file. CI sets COMPOSE_FILE=docker-compose.yml explicitly so
these are excluded from test runs, reducing the number of images to
pull from 14 to 10. Local docker compose auto-loads both files so dev
workflow is unchanged.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add x-build YAML anchor and apply it to all services using the
appwrite-dev image so docker compose pull --ignore-buildable correctly
skips them instead of trying to pull from Docker Hub.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Updated docker-compose.yml to use a pre-built PostgreSQL image instead of a custom build.
- Removed obsolete PostgreSQL Dockerfiles from the project.
- Updated composer.lock with a new content hash.
Jake BarnbyandClaude Opus 4.7
Matej Bačo
Chirag Aggarwal
ArnabChatterjee20k
ArnabChatterjee20k
loks0nandClaude Opus 4.6