Data Validator for VS Code

An AI-powered linter, live verifier, and auto-remediator for your entire engineering ecosystem. Built with pluggable rule packs, this extension acts as a universal validation platform for Airflow DAGs, dbt models, SQL migrations, CI/CD pipelines, Kubernetes manifests, and more. It catches structural errors before they are deployed, verifies external references, and suggests one-click AI fixes right in your editor.

Key Features

1. Universal Validation Platform

The core engine supports hot-swappable Rule Packs allowing it to validate multiple ecosystems natively inside VS Code:

• Apache Airflow DAGs: Detects circular dependencies (>>), unresolved imports, and deprecated variables (e.g., schedule_interval).
• dbt Models (Coming Soon): Validate model references, macros, and schema configurations.
• SQL Migrations: Finds duplicate columns and syntax errors in inline queries or standalone scripts using sqlglot.
• Infrastructure (Coming Soon): CI/CD (GitHub Actions) and Kubernetes manifest linting.

2. Live Target Verification (Data Warehouses)

Extracts source tables and columns from the DAG's SQL queries and verifies them against your live databases.

• Supported Warehouses: Doris, BigQuery, Snowflake.
• Fuzzy Matching: If you misspell a column name, the extension suggests the closest match from the live warehouse.
• Airflow API: Connects to your Airflow REST API to check if referenced connection IDs actually exist in the target environment.

3. AI-Powered Auto-Remediation

Powered by Google Gemini (default) or Anthropic Claude.

• Gated Fixes: When you encounter an error, click "Fix with AI" in the validator dashboard. The AI will stream code fix variants (low, medium, high impact).
• Deterministic Safety: The extension runs the proposed AI changes through the static-check engine in the background. If the AI introduces new errors, the fix is gated to prevent you from applying broken code.

Installation

1. Install the extension from the VS Code Marketplace.
2. Ensure you have Python 3.10+ installed on your system.
3. Install the required Python backend engine packages in your active environment:

   pip install sqlglot google-generativeai pymysql google-cloud-bigquery snowflake-connector-python
   

Configuration

Setting your AI Provider (Gemini / Claude)

The extension securely stores API keys in your OS keychain.

1. Open the Command Palette (Ctrl+Shift+P or Cmd+Shift+P).
2. Run AI Validator: Set API Key.
3. Select your provider (gemini or claude) and paste your API key.

Configuring Live Data Connections

To enable live checks against your warehouse or Airflow instance, configure your connections in the VS Code settings.

1. Go to Settings (Ctrl+,).
2. Search for AI Validator Connections.
3. Add a connection profile (e.g., Doris, Snowflake, or Airflow).
4. Run AI Validator: Set Connection Password from the Command Palette to securely save the password/token for that connection profile.

How to Use

1. Open a DAG: Simply open any Python file containing an Airflow DAG.
2. View Lints: You will immediately see inline squiggles for syntax errors, circular dependencies, or deprecated calls. Hover over them for details.
3. Open Dashboard: Run AI Validator: Open Validator Panel from the Command Palette to see a comprehensive overview of all passed and failed checks for your current DAG.
4. Apply Fixes: Click Fix with AI in the panel to stream structural code fixes directly into the side-by-side diff viewer. You can accept the changes with a single click.

Security & Privacy

• Local First: All static analysis, DAG parsing, and engine rules run entirely locally on your machine via a spawned Python process.
• Secure Credentials: All passwords and API keys are stored securely in the VS Code SecretStorage API (backed by Windows Credential Manager, macOS Keychain, or Linux Secret Service).
• AI Redaction: Before sending code to the AI provider, the engine automatically redacts sensitive connection tokens and passwords.