refactor: simplify env config to a local .env

.env.example

@@ -0,0 +1,19 @@
+# Public (non-secret) runtime config. Expo inlines these into the app at build
+# time, and ONLY variables prefixed EXPO_PUBLIC_ are exposed to the client.
+# See https://docs.expo.dev/guides/environment-variables/
+#
+# `yarn setup:oss` / `yarn setup:artsy` copy this to `.env` (gitignored) if you
+# don't have one. Fill in real values as needed — every var is optional and the
+# app boots fine with them left blank (Unleash + Sentry stay disabled).
+#
+# NOTE: secrets (e.g. ARTSY_API_CLIENT_KEY / ARTSY_API_CLIENT_SECRET) are NOT
+# env vars — they are read through react-native-keys from keys.development.json
+# / keys.production.json. See docs/configuration.
+
+# Unleash feature flags — leave blank to disable remote flags (all flags resolve
+# to false and no network request is made).
+EXPO_PUBLIC_UNLEASH_URL=
+EXPO_PUBLIC_UNLEASH_CLIENT_KEY=
+
+# Sentry error reporting — leave blank to disable.
+EXPO_PUBLIC_SENTRY_DSN=

docs/environment_variables_configuration.md

@@ -1,12 +1,31 @@
 ## Environment Variables Configuration
 
-Energy uses [shared configuration](https://github.com/artsy/README/blob/main/playbooks/development-environments.md#shared-configuration) to distribute common and sensitive configuration values. The [setup script](../scripts/prepare-env-vars) will download `.env.shared` from S3. Then it will add the environment variables ovverrides and any developer custom configuration from `.env.example` to `.env`. Finally, it will merge both into a `.env.dev` file that will be used for both builds.
+Public runtime config lives in a local **`.env`** file at the project root. Expo
+inlines it at build time, and **only** variables prefixed `EXPO_PUBLIC_` are
+exposed to the app — see the
+[Expo environment variables guide](https://docs.expo.dev/guides/environment-variables/).
 
-> **Note**: Whenever you change an environment variable, you will need to rebuild the app again
+`yarn setup:oss` and `yarn setup:artsy` copy the tracked **`.env.example`** to
+`.env` (gitignored) if you don't have one. You can also do it by hand:
 
-## Add/Update remote environment variable
-
-```
-aws s3 cp s3://artsy-citadel/dev/.env.energy .env.shared
-aws s3 cp .env.shared s3://artsy-citadel/dev/.env.energy
+```sh
+cp .env.example .env
 ```
+
+Every variable is optional — the app boots fine with them left blank:
+
+| Variable | Purpose | Blank behavior |
+| --- | --- | --- |
+| `EXPO_PUBLIC_UNLEASH_URL` | Unleash proxy URL | Remote flags disabled (all flags `false`, no network) |
+| `EXPO_PUBLIC_UNLEASH_CLIENT_KEY` | Unleash client key | Remote flags disabled |
+| `EXPO_PUBLIC_SENTRY_DSN` | Sentry DSN | Error reporting disabled |
+
+> **Note:** Expo inlines env vars at **build time**, so after changing a value
+> you must rebuild the app.
+
+## Secrets are not env vars
+
+API secrets (e.g. `ARTSY_API_CLIENT_KEY` / `ARTSY_API_CLIENT_SECRET`) are read
+through [`react-native-keys`](https://github.com/numandev1/react-native-keys)
+from `keys.development.json` / `keys.production.json`, **not** from `.env`. See
+[Configuration](./configuration).

scripts/setup-env-for-artsy

@@ -1,11 +1,18 @@
 #!/usr/bin/env bash
 set -euxo pipefail
 
-# Resolve the sibling scripts by this file's own location and run them from the
-# project root, so `yarn setup:artsy` works from the repo root (where yarn runs
-# it) and not only from inside scripts/.
+# Resolve sibling scripts by this file's own location and run from the project
+# root, so `yarn setup:artsy` works from the repo root (where yarn runs it) and
+# not only from inside scripts/.
 SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
 cd "$SCRIPT_DIR/.."
 
+# Download Artsy's private fonts from S3 (needs Artsy AWS access).
 "$SCRIPT_DIR/download-fonts"
-"$SCRIPT_DIR/prepare-env-vars"
+
+# Seed .env from the tracked example if you don't have one. Expo auto-loads
+# .env and inlines EXPO_PUBLIC_* at build time; fill in real values by hand.
+# (Secrets live in keys.*.json via react-native-keys, not here.)
+if [ ! -f .env ]; then
+  cp .env.example .env
+fi

scripts/setup-oss

@@ -51,6 +51,14 @@ for env in development production; do
   fi
 done
 
+echo "▸ .env (Expo public config)"
+if [ -f .env ]; then
+  echo "  • keeping existing .env"
+else
+  cp .env.example .env
+  echo "  • created .env from .env.example — fill in EXPO_PUBLIC_* values (optional)"
+fi
+
 cat <<'EOF'
 
 ✅ OSS setup complete. Next: