fix: close backup review findings

.superpowers/sdd/final-review-fix-report.md

@@ -0,0 +1,37 @@
+# Backup Final Review Fix Report
+
+## Implemented findings
+
+- The backup container now validates that `RESTIC_REPOSITORY` resolves strictly
+  below `/backup`, verifies that the Restic repository already has a readable
+  configuration before a scheduled backup, and gives the explicit first-run
+  initialization command when that preflight fails.
+- The backup manifest now records every successfully applied Prisma migration
+  as a JSON object with its name and UTC completion timestamp. The `psql`
+  query uses the existing `DATABASE_URL` connection configuration and stops on
+  query errors.
+- Restore now requires `BACKUP_REPOSITORY`, validates that it resolves strictly
+  below `/backup`, and validates the configured backup mount (including the
+  existing writable probe) before `restore-live` can stop services or replace
+  live data.
+- The backup runbook now documents explicit repository initialization, monthly
+  full-read Restic checks, a disposable restore/checksum rehearsal, and
+  post-restore Compose status/log checks.
+
+## Verification
+
+- `bash -n scripts/backup/container-entrypoint.sh`
+- `bash -n scripts/backup/restore.sh`
+- Focused `rg` assertions for repository validation/preflight, migration
+  timestamp metadata, restore mount validation, initialization, maintenance,
+  and post-restore runbook commands.
+- `npx prisma validate`
+- `git diff --check`
+
+## Scope and concerns
+
+- No Docker, NAS, systemd, Restic repository, or live restore was run, per the
+  bounded review scope. The command-level behavior is therefore statically
+  validated only.
+- Existing durable STL handling, the exact live-restore confirmation flag, and
+  rollback/safety-artifact behavior were retained.

scripts/backup/README.md

@@ -113,7 +113,20 @@ the Synology share. `BACKUP_APP_VERSION` is optional metadata. The production
 Compose environment must also retain its existing database and application
 secrets; do not add any secrets to Git.
 
-## 3. Install the nightly systemd job
+## 3. Initialize the Restic repository once
+
+Before enabling the scheduled job, initialize a new repository explicitly:
+
+```bash
+docker compose --profile backup run --rm backup init
+```
+
+The backup container preflight refuses to create a repository implicitly. If a
+scheduled backup reports that the repository is unavailable or uninitialized,
+verify the NFS mount, `BACKUP_REPOSITORY`, and password file before running the
+explicit initialization command for an intended new repository.
+
+## 4. Install the nightly systemd job
 
 The supplied unit assumes the production Compose checkout is
 `/opt/stacks/DragonsStash`. If your deployment lives elsewhere, update the
@@ -136,7 +149,7 @@ catches up after downtime. The first run can take a long time because it uploads
 all existing STL and session data. Later Restic snapshots deduplicate unchanged
 data.
 
-## 4. Monitor and maintain backups
+## 5. Monitor and maintain backups
 
 Inspect the next scheduled run and the last service result:
 
@@ -167,7 +180,29 @@ Choose `BACKUP_RETENTION_DAYS` based on storage capacity and the recovery
 window you need. Watch Synology capacity and investigate any failed timer or
 service promptly.
 
-## 5. Restore modes
+At least monthly, perform a full repository read check:
+
+```bash
+docker compose --profile backup run --rm backup check --read-data
+```
+
+Also rehearse recovery using a disposable staging directory and verify the
+database dump checksum before deleting the rehearsal directory:
+
+```bash
+REHEARSAL_DIR=/var/lib/dragons-stash/backup-staging/monthly-rehearsal-SNAPSHOT_ID
+./scripts/backup/restore.sh restore-to-staging SNAPSHOT_ID "$REHEARSAL_DIR"
+(
+  cd "$REHEARSAL_DIR"/staging/backup-*
+  sha256sum --check manifest/database.dump.sha256
+)
+```
+
+The staging restore also verifies the custom PostgreSQL dump and required data
+directories. Inspect the restored data as appropriate, then remove the
+disposable directory using your approved host cleanup procedure.
+
+## 6. Restore modes
 
 Run restore commands from the production Compose checkout after loading the
 same backup environment used by systemd (for example, as root with
@@ -190,6 +225,14 @@ safety artifacts and staging directory, and attempts rollback after replacement
 has begun. Review the reported paths and service health before manually
 starting services.
 
+After a successful live restore, confirm the services are running and inspect
+their recent logs before declaring recovery complete:
+
+```bash
+docker compose ps
+docker compose logs --tail=100 app worker bot
+```
+
 For a non-destructive recovery rehearsal, choose a snapshot from `list` and
 restore it to a fresh staging directory, for example:
 

scripts/backup/container-entrypoint.sh

@@ -27,6 +27,40 @@ require_directory() {
   fi
 }
 
+validate_restic_repository() {
+  local canonical_repository
+
+  canonical_repository="$(realpath -m -- "$RESTIC_REPOSITORY")"
+  if [[ "$canonical_repository" == "$BACKUP_ROOT" || "$canonical_repository" != "$BACKUP_ROOT"/* ]]; then
+    printf 'RESTIC_REPOSITORY must be strictly below /backup; got %s.\n' "$RESTIC_REPOSITORY" >&2
+    return 1
+  fi
+
+  RESTIC_REPOSITORY="$canonical_repository"
+  export RESTIC_REPOSITORY
+}
+
+validate_restic_configuration() {
+  require_value RESTIC_REPOSITORY
+  require_value RESTIC_PASSWORD_FILE
+
+  if [[ ! -r "$RESTIC_PASSWORD_FILE" ]]; then
+    printf 'Restic password file %s is not readable.\n' "$RESTIC_PASSWORD_FILE" >&2
+    return 1
+  fi
+
+  require_directory "$BACKUP_ROOT"
+  validate_restic_repository
+}
+
+ensure_repository_initialized() {
+  if ! restic cat config >/dev/null; then
+    printf 'Restic repository %s is not initialized or could not be opened. Verify the backup mount, repository path, and password.\n' "$RESTIC_REPOSITORY" >&2
+    printf 'For an intended first-time repository, initialize it explicitly with: docker compose --profile backup run --rm backup init\n' >&2
+    return 1
+  fi
+}
+
 json_escape() {
   local input="$1"
   local output=""
@@ -72,22 +106,16 @@ run_backup() {
   local timestamp
   local checksum
   local app_version
+  local applied_migrations
 
   require_value DATABASE_URL
-  require_value RESTIC_REPOSITORY
-  require_value RESTIC_PASSWORD_FILE
   require_value BACKUP_RETENTION_DAYS
-
+  validate_restic_configuration
-  if [[ ! -r "$RESTIC_PASSWORD_FILE" ]]; then
-    printf 'Restic password file %s is not readable.\n' "$RESTIC_PASSWORD_FILE" >&2
-    return 1
-  fi
-
-  require_directory "$BACKUP_ROOT"
   require_directory "$STAGING_ROOT"
   require_directory "$UPLOADS_PATH"
   require_directory "$TDLIB_WORKER_PATH"
   require_directory "$TDLIB_BOT_PATH"
+  ensure_repository_initialized
 
   timestamp="$(date -u +'%Y-%m-%dT%H:%M:%SZ')"
   RUN_DIR="$STAGING_ROOT/backup-${timestamp//[:]/}-$$"
@@ -105,6 +133,26 @@ run_backup() {
   sha256sum "$RUN_DIR/database.dump" > "$RUN_DIR/manifest/database.dump.sha256"
   checksum="$(awk '{print $1}' "$RUN_DIR/manifest/database.dump.sha256")"
   app_version="${BACKUP_APP_VERSION:-unknown}"
+  applied_migrations="$(
+    psql --dbname "$DATABASE_URL" --no-psqlrc --tuples-only --no-align --quiet \
+      --set ON_ERROR_STOP=on <<'SQL'
+      SELECT COALESCE(
+        json_agg(
+          json_build_object(
+            'name', "migration_name",
+            'finishedAtUtc', to_char(
+              "finished_at" AT TIME ZONE 'UTC',
+              'YYYY-MM-DD"T"HH24:MI:SS.US"Z"'
+            )
+          )
+          ORDER BY "finished_at", "migration_name"
+        ),
+        '[]'::json
+      )::text
+      FROM "_prisma_migrations"
+      WHERE "finished_at" IS NOT NULL AND "rolled_back_at" IS NULL;
+SQL
+  )"
 
   cat > "$RUN_DIR/manifest/backup-manifest.json" <<EOF
 {
@@ -112,6 +160,7 @@ run_backup() {
   "repository": "$(json_escape "$RESTIC_REPOSITORY")",
   "retentionDays": "$(json_escape "$BACKUP_RETENTION_DAYS")",
   "applicationVersion": "$(json_escape "$app_version")",
+  "appliedPrismaMigrations": $applied_migrations,
   "databaseDump": {
     "filename": "database.dump",
     "sha256": "$(json_escape "$checksum")"
@@ -142,11 +191,17 @@ EOF
 
 run_restore() {
   shift
+  validate_restic_configuration
   exec restic restore "$@"
 }
 
+run_restic() {
+  validate_restic_configuration
+  exec restic "$@"
+}
+
 case "${1:-backup}" in
   backup) run_backup ;;
   restore) run_restore "$@" ;;
-  *) exec restic "$@" ;;
+  *) run_restic "$@" ;;
 esac

scripts/backup/restore.sh

@@ -3,6 +3,7 @@ set -Eeuo pipefail
 
 readonly CONFIRM_REPLACE_LIVE_DATA="--confirm-replace-live-data"
 readonly STAGING_CONTAINER_ROOT="/staging"
+readonly BACKUP_CONTAINER_ROOT="/backup"
 readonly -a LIVE_SERVICES=(app worker bot)
 
 RESTORED_DUMP=""
@@ -40,10 +41,42 @@ validate_environment() {
   require_value BACKUP_MOUNT_PATH
   require_value BACKUP_STAGING_PATH
   require_value BACKUP_RESTIC_PASSWORD_FILE
+  require_value BACKUP_REPOSITORY
   if [[ ! -r "$BACKUP_RESTIC_PASSWORD_FILE" ]]; then
     printf 'Restic password file %s is not readable.\n' "$BACKUP_RESTIC_PASSWORD_FILE" >&2
     return 1
   fi
+  validate_backup_repository
+}
+
+validate_backup_repository() {
+  local repository="$BACKUP_REPOSITORY"
+  local canonical_repository
+
+  canonical_repository="$(realpath -ms -- "$repository")"
+  if [[ "$canonical_repository" == "$BACKUP_CONTAINER_ROOT" || "$canonical_repository" != "$BACKUP_CONTAINER_ROOT"/* ]]; then
+    printf 'BACKUP_REPOSITORY must be strictly below /backup; got %s.\n' "$repository" >&2
+    return 1
+  fi
+}
+
+validate_backup_mount() {
+  local probe_file
+
+  if ! mountpoint --q "$BACKUP_MOUNT_PATH"; then
+    printf 'Backup mount %s is not an active mountpoint.\n' "$BACKUP_MOUNT_PATH" >&2
+    return 1
+  fi
+
+  if ! probe_file="$(mktemp "$BACKUP_MOUNT_PATH/.dragons-stash-backup-write-probe.XXXXXX")"; then
+    printf 'Backup mount %s is not writable.\n' "$BACKUP_MOUNT_PATH" >&2
+    return 1
+  fi
+
+  if ! rm -f -- "$probe_file"; then
+    printf 'Unable to remove writable probe %s.\n' "$probe_file" >&2
+    return 1
+  fi
 }
 
 backup_restic() {
@@ -352,6 +385,7 @@ restore_live() {
   local timestamp
   local container_staging_dir
   validate_environment
+  validate_backup_mount
   docker compose --profile backup config --quiet
   project_name="$(compose_project_name)"
   verify_snapshot "$snapshot_id"

scripts/backup/run-backup.sh

@@ -2,6 +2,7 @@
 set -Eeuo pipefail
 
 readonly LOCK_FILE="/run/lock/dragons-stash-backup.lock"
+readonly BACKUP_CONTAINER_ROOT="/backup"
 readonly -a MANAGED_SERVICES=(app worker bot)
 
 declare -a RUNNING_SERVICES=()
@@ -22,6 +23,17 @@ validate_environment() {
   require_value BACKUP_RETENTION_DAYS
 }
 
+validate_backup_repository() {
+  local repository="${BACKUP_REPOSITORY:-/backup/restic}"
+  local canonical_repository
+
+  canonical_repository="$(realpath -ms -- "$repository")"
+  if [[ "$canonical_repository" == "$BACKUP_CONTAINER_ROOT" || "$canonical_repository" != "$BACKUP_CONTAINER_ROOT"/* ]]; then
+    printf 'BACKUP_REPOSITORY must be strictly below /backup; got %s.\n' "$repository" >&2
+    return 1
+  fi
+}
+
 validate_backup_mount() {
   local probe_file
 
@@ -73,6 +85,7 @@ restart_running_services() {
 
 main() {
   validate_environment
+  validate_backup_repository
 
   exec 9>"$LOCK_FILE"
   if ! flock -n 9; then