fix: validate acceptedAssets to prevent silent malformed input (#9)

docs/queue-dead-letter-maintenance.md

@@ -1,18 +1,21 @@
 # Queue dead-letter maintenance — changes made
 
 Summary
+
 - Bounded failed-job retention: default job options updated so failed jobs are retained only up to a bounded count instead of forever.
 - Daily maintenance cron: a scheduled service prunes old failed jobs, persists a write-only record in Postgres, optionally reports to Sentry, and removes old failed jobs from Redis.
 - Observability: a Prometheus gauge `bull_dead_letter_count{queue="..."}` is updated with the current failed-job count per queue.
 - Persistence: added a Prisma model `DeadLetter` (mapped to table `dead_letters`) to store pruned failure records for audit.
 
 Files added/modified
+
 - Modified: [src/queue/queue.module.ts](src/queue/queue.module.ts) — set `removeOnFail: 1000`, registered `ScheduleModule`, and added the maintenance service
 - Added: [src/queue/queue-maintenance.service.ts](src/queue/queue-maintenance.service.ts) — maintenance cron, Sentry integration, Prometheus metric
 - Modified: [prisma/schema.prisma](prisma/schema.prisma) — added `DeadLetter` model
 - Modified: [package.json](package.json) — added `@nestjs/schedule` and `prom-client` dependencies
 
 Behavior details
+
 - Default job options (applied to the main queues: `email`, `contract-events`, `analytics`, `export`):
   - `attempts: 3`
   - `backoff: { type: 'exponential', delay: 5000 }`
@@ -28,30 +31,40 @@ Behavior details
   - Updates Prometheus gauge `bull_dead_letter_count` for each queue via `prom-client`.
 
 Deployment & local steps
+
 1. Install dependencies:
+
 ```bash
 npm install
 ```
+
 2. Generate Prisma client and run migrations (create migration to add `dead_letters`):
+
 ```bash
 npx prisma generate
 npx prisma migrate dev --name add_dead_letters
 ```
+
 3. Set optional environment variables:
-  - `SENTRY_DSN` to enable Sentry reporting.
-  - `DATABASE_URL` for Postgres.
+
+- `SENTRY_DSN` to enable Sentry reporting.
+- `DATABASE_URL` for Postgres.
+
 4. Build and run:
+
 ```bash
 npm run build
 npm start:dev
 ```
 
 Notes & follow-ups
+
 - I did not add an HTTP metrics endpoint — if you want `/metrics` exported for Prometheus scraping I can add a small controller that exposes `prom-client` metrics.
 - The branch with these changes is: `feat/queue-dead-letter-maintenance`.
 - Open PR URL: https://github.com/coderolisa/OrbitChain-API/pull/new/feat/queue-dead-letter-maintenance
 
 If you'd like, I can also:
+
 - Add an HTTP `/metrics` endpoint.
 - Make the prune retention and retained-fail-count configurable via environment variables.
 - Add unit tests around the maintenance service.

eslint.config.mjs

@@ -48,7 +48,7 @@ export default tseslint.config(
       '@typescript-eslint/require-await': 'warn',
       '@typescript-eslint/restrict-template-expressions': 'warn',
       '@typescript-eslint/prefer-as-const': 'warn',
-      "prettier/prettier": ["error", { endOfLine: "lf" }],
+      'prettier/prettier': ['error', { endOfLine: 'lf' }],
     },
   },
 );