# Stitch Data MCP for AI Agents AI Agent Connect > Stitch Data MCP lets you manage your entire data pipeline from your favorite AI client. You can create sources, push records to your warehouse, and check sync statuses without leaving your IDE. It connects your Stitch account to your agent so you can orchestrate ETL workflows using plain English. ## Overview - **Category:** data-analytics - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_mABPCZrnZTeawQZOSWTl9iyadYUD2LtJD9LLWWGw/ai-agent-connect - **Tags:** etl, data-pipeline, data-ingestion, warehouse, stitch-data ## Description This Connector connects your Stitch Data account directly to your AI agent so you can manage your data infrastructure through a conversation. Instead of jumping between the Stitch dashboard and your terminal, you can tell your agent to set up new destinations like Snowflake or S3. You can also push raw data batches to the Import API while the agent handles the schema validation for you. It is a way to keep your warehouse up to date by triggering replication jobs or checking extraction logs on the fly. Because Vinkius makes it easy to connect this to your existing workflow, you can handle account administration and session management without hunting for API keys. Whether you are a data engineer debugging a failed load or an analyst pushing a custom batch for a one-off report, this Connector puts the controls for your data ingestion right where you are already working. ## Tools ### create_ephemeral_session Generate a temporary token for the Connect JavaScript client. This keeps your session secure. ### delete_destination Remove a destination from your configured list. Use this to clean up old warehouse targets. ### create_destination Configure a new warehouse destination for your account. This helps you set up where your data lands. ### create_source Create a new data source to bring in external information. Use this to add new apps to your pipeline. ### delete_source Delete a source you no longer need. This removes the connection from your Stitch account. ### get_import_status Check if the Stitch Import API is currently operating correctly. It is a quick way to verify system health. ### list_destination_types See which warehouse types like Snowflake or S3 are available. This helps you choose the right target. ### list_destinations View all the destinations you have already configured. This gives you a quick overview of your targets. ### list_extractions See a list of the most recent extraction jobs. Use this to track progress on your data pulls. ### list_loads View the history of recent data loading attempts. This is perfect for debugging failed imports. ### list_source_types See all the different types of sources you can connect. Use this to explore what data you can bring in. ### list_sources See every data source currently linked to your account. This helps you audit your active connections. ### list_streams See the tables and streams available within a specific source. Use this to see what data is ready to sync. ### push_import_batch Push a batch of records for a single table to the Import API. This is best for keeping your data types consistent. ### push_import_data Push data for one or more tables to the Import API without schema enforcement. Use this for rapid data loading. ### start_sync Manually trigger a replication job for your data. Use this to force an update on a specific source. ### update_destination Change the settings for an existing destination. Use this to modify your warehouse configurations. ### update_source Update, pause, or unpause a specific data source. This gives you quick control over your active streams. ### update_stream_metadata Select specific streams and fields for your replication. This lets you control exactly what data gets synced. ### validate_import_data Test your credentials and formatting without saving any data. This is a safe way to check your setup. ### create_account Create a new Stitch client account using partner credentials. This is for partner-level account management. ## Prompt Examples **Prompt:** ``` List my active sources. ``` **Response:** ``` I've checked your Stitch account. You have 4 active sources: * **Salesforce** (ID: src_123) * **Shopify** (ID: src_456) * **Zendesk** (ID: src_789) * **S3 Bucket** (ID: src_000) ``` **Prompt:** ``` Can you check why my last load failed? ``` **Response:** ``` I've looked at your recent activity. Your last load attempt for the **orders** table failed at 10:14 AM. **Error Details:** * **Status:** Failed * **Reason:** Timeout error on the source side * **Source:** Shopify You might want to check if the Shopify API is currently experiencing downtime. ``` **Prompt:** ``` Push these 50 user records to the production table. ``` **Response:** ``` I'm pushing those records now. I've used the batch import tool to send them to your production table. They are now queued for ingestion into your warehouse. ``` ## Capabilities ### Create new data sources Add new connections to your pipeline without manual configuration. ### Push data batches Send raw records to your warehouse with built-in schema validation. ### Manage destinations Configure and update warehouse targets like Redshift or Snowflake. ### Monitor sync status Check extraction logs and load statuses to find where pipelines stall. ### Trigger replication Manually start sync jobs to ensure your data is current. ### Validate credentials Test your import settings before you actually move any records. ## Use Cases ### Debugging a failed sync An engineer asks the agent to check list_loads and list_extractions to see why a Redshift load failed during the nightly run. ### Ad-hoc data push An analyst wants to push a CSV of new user records to a table and asks the agent to use push_import_data to get them in quickly. ### Quick source audit A manager asks the agent to list_sources to see which third-party apps are currently connected to the Stitch account. ### Setting up a new warehouse A developer asks the agent to list_destination_types and then create_destination to set up a new Snowflake environment. ## Benefits - Stop manual clicking: Use list_sources and list_destinations to see your whole setup in one glance without opening the dashboard. - Faster data loading: Push records instantly using push_import_batch for quick testing and validation of your data types. - Easier debugging: Check list_extractions and list_loads to find errors and see why a sync failed without hunting for logs. - Dynamic configuration: Update sources with update_source or update_stream_metadata using natural language to change your sync logic. - Secure testing: Use validate_import_data to check your credentials and formatting before you actually move any records to your warehouse. ## How It Works The bottom line is you get a voice-controlled dashboard for your entire ETL pipeline. 1. Subscribe to the Stitch Data MCP on Vinkius. 2. Add your Stitch Connect Token, Import Token, and Region to your AI client. 3. Ask your agent to list sources, push data, or check logs. ## Frequently Asked Questions **Can I use Stitch Data MCP to manage my data sources?** Yes, you can list, create, and update all your data sources through your AI client. This allows you to manage your connections without leaving your workspace. **How does this help with data ingestion?** It lets you push data batches directly to your warehouse and check the status of those imports in real time. You can handle both manual pushes and automated syncs. **Can I use this to check my sync status?** Yes, you can ask your agent to show you the latest extraction logs or load attempts to see if your data is moving. It provides a quick way to debug pipeline issues. **Is it safe to test my credentials?** You can use the validation tool to check your import settings without actually saving any data to your warehouse. It's a great way to verify your setup safely. **Can I trigger a manual sync?** Yes, you can tell your agent to start a replication job for any of your connected sources immediately. This is helpful for forcing updates on time-sensitive data. **Does this work with Snowflake and Redshift?** Yes, it supports managing destinations for common warehouses like Snowflake, Redshift, and S3. You can configure and update these targets via your agent. **How can I verify if my data format is correct before pushing it to the warehouse?** You can use the `validate_import_data` tool. It functions exactly like a data push but only tests your credentials and data formatting without actually persisting any records. **Is there a way to check if the Stitch Import API is currently operational?** Yes, use the `get_import_status` tool. It checks the operational status of the Import API and returns whether it is functioning correctly. **How do I see which types of sources I can connect to my Stitch account?** Run the `list_source_types` tool. It will provide a comprehensive list of all available source connectors supported by Stitch.