Can I vibecode Parabola?

Build a deliberately narrow, functional personal substitute for Parabola, focusing on reliable tabular data transformation rather than a sprawling visual canvas.
Use exactly this stack: Next.js 15 (App Router) + TypeScript + PostgreSQL + BullMQ.

Primary job: Build one explicit data automation pipeline (e.g., Extract from CSV/API -> Transform -> Load to DB/API) with a trigger, validated sequential steps, retries, logs, and a manual replay button.
Start from an empty folder and create the complete working project. Make the default mode single-user and private.

Feature Requirements:
1. Pipeline Architecture: Define data workflows in code as a series of typed transformation steps (e.g., filter rows, map columns, aggregate data). The engine should pass datasets (as arrays of objects or dataframes) between these steps.
2. Job Execution: Use BullMQ to handle pipeline execution asynchronously. This ensures large datasets don't block the UI and allows for scheduling (cron).
3. Error Handling & State: Implement robust retries for network-dependent steps (like API fetching). Store the intermediate state of the data between steps in PostgreSQL so that if a pipeline fails at step 4, it can be resumed from step 4 without re-fetching data.
4. Dashboard: Build a Next.js UI that lists available pipelines, shows run history, and displays a step-by-step breakdown of a specific run (showing row counts and data samples at each step). Include a "Replay Pipeline" button.
5. Export/Import: Ensure the final transformed data can easily be exported as CSV or JSON.

Constraints & Scope:
- Store user data locally unless the core job requires the declared self-hosted database.
- Do not add analytics, telemetry, ads, or third-party accounts. Put every secret in `.env` and provide `.env.example`.
- Implement a polished, accessible interface with clear empty, loading, validation, success, and failure states.
- Validate untrusted input and never log secrets.
- Deliberately exclude paid-product advantages: a drag-and-drop DAG builder, durable execution at massive enterprise scale, schema drift anomaly detection, and hundreds of maintained API connectors.

Deliverables:
- Complete source code.
- Write focused unit tests for the core data transformation functions (e.g., joining, filtering).
- Add one end-to-end smoke test that processes a sample dataset through a mock pipeline.
- Create a README with setup, permissions, architecture, and backup instructions.
- Add standard scripts (install, dev, test, build) and fix all errors before finishing.