env fix

.dockerignore

@@ -1,16 +1,15 @@
-# What does not reach the image.
+# Nothing in here reaches the image.
 #
-# `.env.production` IS copied in — see the Dockerfile's runtime stage. The
+# `.env*` most of all: the Dockerfile does `COPY . .`, and the production
-# container reads its own configuration from that file, so the deployment does
+# credentials in `.env.production` were being baked into every image built
-# not depend on every variable having been typed into a hosting platform's
+# from this folder. The running container gets its environment from the
-# settings screen. Anything the platform DOES set still wins: godotenv never
+# platform (Dokploy / Kubernetes), never from a file.
-# overwrites a variable that is already present in the environment.
 #
-# Every other `.env.*` stays out. `.env.local` and `.env.secrets` are one
+# An exception for `.env.production` was added on 2026-09-25 so the container
-# developer's machine, and `.env.secrets` in particular is the file that holds
+# could read its own configuration, and the deploy came back 502 on every
-# a key — it must never be inside an image.
+# endpoint. Reverted. The committed file is a stale snapshot; letting it fill
+# whatever the platform leaves unset is not a safe default.
 .env*
-!.env.production
 
 .git
 .claude

Dockerfile

@@ -28,22 +28,26 @@ WORKDIR /app
 COPY --from=builder /app/server /app
 COPY nearle-gear-firebase-adminsdk-l9oha-23ca3b3609.json .
 
-# The container's own configuration.
+# Nearle Buddy's credential, as ONE container variable.
 #
-# Copied in so the deployment does not depend on every variable having been
+# Not an env file. `COPY .env.production .` was tried on 2026-09-25 and took the
-# entered into a hosting platform's settings screen. A variable missed there is
+# backend down with 502 on every endpoint: that file declares twenty-three
-# a variable unset in production, and the failure is silent — the assistant sat
+# variables, and godotenv fills any the platform leaves unset, so a stale
-# switched off for two days for exactly that reason, with nothing on any screen
+# committed DB or Redis value replaced a live one and the process died at boot.
-# saying which value was absent.
+# Twenty-three variables shipped to deliver one.
 #
-# Anything the platform DOES set still wins: `godotenv` only fills variables
+# A single ENV cannot do that — it sets this name and no other. A value set on
-# that are not already in the environment, so Dokploy can override any line in
+# the platform still wins, because `docker run -e` overrides a Dockerfile ENV,
-# this file without the file having to change.
+# so this is a default rather than an override.
-COPY .env.production .
+#
+# Provider, endpoint and model are constants in config.go, so this is the only
+# thing the assistant needs to come up.
+ENV ASSISTANT_API_KEY=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
 
-# Decides which rules config.Load applies — production insists on a signing
+# No `.env.*` is copied in (see .dockerignore), so this only decides which rules
-# secret and never falls back to localhost values — and which file above is
+# config.Load applies: production insists on a signing secret and never falls
-# read: loadEnvFiles reads `.env.<APP_ENV>` then `.env`.
+# back to localhost values. Every other real value comes from the platform's
+# environment settings, exactly as before.
 ENV APP_ENV=production
 
 # Must match APP_PORT in the platform's environment (1009 in production).