# Airbyte MCP for AI Agents MCP > Airbyte MCP lets your AI agent monitor entire data integration pipelines. Check sync job status, list all active sources (like Postgres or Stripe), and audit configured destinations (such as Snowflake or BigQuery) instantly via conversation. It keeps your modern data stack running without you touching a dashboard. ## Overview - **Category:** brain-trust - **Price:** Free - **Tags:** etl-pipelines, data-integration, data-warehousing, pipeline-monitoring, data-sync, data-engineering ## Description Your AI agent can talk directly to your Airbyte instance, giving you conversational visibility into every part of your ETL/ELT process. Instead of logging into the dashboard and clicking through pages just to see if everything ran overnight, your agent handles the audit automatically. You tell it what you need—like checking yesterday's Postgres sync rate or listing all destinations pointing to Snowflake—and get a clean answer back immediately. It’s like having a dedicated data ops engineer on standby 24/7. This MCP connects that oversight capability directly into Vinkius, making your whole data flow visible through any compatible AI client. ## Tools ### list_sources Lists every available source system connected to Airbyte. ### get_source Retrieves the detailed configuration for a single data source. ### list_destinations Lists all configured target destinations where data is sent. ### list_connections Lists every active pipeline connection between sources and destinations. ### get_connection Gets the specific details for one established data sync connection. ### list_jobs Lists job history, showing when a connection last ran and whether it succeeded or failed. ### list_workspaces Retrieves the list of separate workspaces within your Airbyte environment. ## Prompt Examples **Prompt:** ``` What were the sync results for my Stripe connection last week? ``` **Response:** ``` **Stripe Connection Job History** * **Date:** 2026-06-12 (Yesterday) * Status: **SUCCESS** * Records Sync'd: 1,450 * Duration: 3m 12s * **Date:** 2026-06-11 (Day Before) * Status: **FAILED** * Reason: Missing authentication token for the billing API. * Records Sync'd: 0 The failure on June 11th was due to a credential issue. You need to update your Stripe credentials in Airbyte. ``` **Prompt:** ``` List all active destinations and check if Snowflake is configured. ``` **Response:** ``` You have **4** active data destinations: * ✅ **Snowflake Data Cloud**: (ID: ...a12) Used for core reporting. * ✅ **Google BigQuery Prod**: (ID: ...b3c) Used for marketing analytics. * 🟡 **Amazon S3 Bucket**: (ID: ...d4e) Used for raw JSON dumps. *Status check needed.* * ❌ **Legacy Redshift Cluster**: (ID: ...f5g) Connection is listed but appears inactive. Snowflake looks good to go, but I recommend checking the status on your S3 bucket. ``` **Prompt:** ``` Show me all data origins and check their configuration IDs. ``` **Response:** ``` **Airbyte Sources Found (Total: 5)** 1. **Postgres DB**: ID `pg-prod-001`. Status: Active. 2. **Stripe API**: ID `stripe-live`. Status: Active, last sync 3 hours ago. 3. **Salesforce CRM**: ID `sf-uat-test`. Status: Warning (Credentials nearing expiry). 4. **Google Sheets**: ID `gsheet-leads`. Status: Active. 5. **Internal API**: ID `internal-users`. Status: Active. ``` ## Capabilities ### List all configured sources Retrieves a full list of every data origin (sources) you've connected in Airbyte. ### Get details for a specific source Pulls detailed configuration and status information for one particular data source. ### List all destinations Provides a comprehensive list of every target warehouse or destination configured in Airbyte. ### List active sync connections Shows all the established data pipelines (connections) that move data from sources to destinations. ### Get connection details Fetches specific details, configuration, and status for a single data synchronization connection. ### Track job history and success rates View historical records of sync jobs for any given connection, detailing success or failure. ### List Airbyte workspaces Retrieves a list of all active workspace environments within your Airbyte account. ## Use Cases ### The nightly Postgres sync failed A data engineer asks their agent, 'What was the status of the Postgres source connection last night?' The agent runs `list_jobs` and tells them exactly which job failed, why it timed out, and when the previous run succeeded. ### I need to audit our warehouse targets An analytics engineer asks, 'Show me every destination we've pointed data towards.' The agent uses `list_destinations` and returns a clean list of all configured endpoints like Snowflake and BigQuery. ### Which sources are currently connected? A manager needs to know what systems feed the data lake. They ask, 'List all active data origins.' The agent uses `list_sources` to provide a clean count and list of everything from Postgres to Stripe. ### Verify connection paths for new projects A team member asks the agent to summarize current pipelines. The agent calls `list_connections`, providing a comprehensive overview of all data movement paths currently running. ## Benefits - Stop checking dashboards manually. Your agent directly queries the job history using `list_jobs` to tell you instantly if a nightly sync failed. - Get a full inventory of your infrastructure by running `list_sources` and `list_destinations`, giving you immediate visibility into all data origins and targets. - Quickly troubleshoot connectivity issues. Use `get_connection` to pull detailed status for a specific pipeline, saving minutes of dashboard clicking. - Understand the whole scope of your setup by calling `list_connections`. You see every active path from source to warehouse at a glance. - Maintain environment oversight by running `list_workspaces`, confirming that all operational environments are correctly configured. ## How It Works The bottom line is your AI can act as a constant monitor, querying Airbyte’s operational state without needing manual dashboard interaction. 1. Subscribe to this MCP and provide your specific Airbyte API URL and API Key. 2. Your AI client runs diagnostic queries, asking for pipeline status or connection details via the exposed tools. 3. The agent returns structured data—like job history or source lists—which it presents back to you in plain language. ## Frequently Asked Questions **How do I check if my data pipeline ran successfully using the Airbyte MCP for AI Agents?** Your agent checks the job history directly. You simply ask it about a connection, and it tells you the status (success/fail) of specific runs, saving you from clicking through dashboards. **Can I use the Airbyte MCP for AI Agents to see all my data sources?** Yes. You can ask the agent to list all your connected data origins (like Postgres or Stripe) instantly. It gives you a clean, comprehensive inventory of everything feeding your data lake. **Does the Airbyte MCP for AI Agents help me find my warehouse endpoints?** Absolutely. You can list all configured destinations—whether it's Snowflake or BigQuery—so you always know exactly where every piece of data is going. **What if I need to debug a failed sync job with the Airbyte MCP for AI Agents?** You tell your agent which connection failed, and it retrieves the detailed job history. It often includes the error reason (like a missing credential) so you know exactly what needs fixing. **Is the Airbyte MCP for AI Agents better than just checking the dashboard?** It's faster and more reliable. Instead of manual clicking, your agent performs automated audits, giving you a summarized report in plain language that highlights exactly what needs attention.