Announcing New Experimental MCP Server inside PyAirbyte

[aaronsteers]

As of v0.29.0, we've added an MCP server directly into PyAirbyte!

https://airbytehq.github.io/PyAirbyte/airbyte/mcp.html

Summary of tools being made available as of this writing:

Connector Management

• list_connectors - List available Airbyte connectors with optional filtering by type (source/destination), install types (python, yaml, java, docker), or keywords
• get_connector_info - Get documentation URL and information for a specific connector
• list_connector_config_secrets - List available config secret names for a given connector
• validate_connector_config - Validate a connector configuration

Source Operations

• list_source_streams - List all streams available in a source connector
• get_source_stream_json_schema - Get the JSON schema for a specific stream in a source connector
• get_stream_previews - Get sample records (previews) from streams in a source connector
• read_source_stream_records - Read records from a specific source stream

Cache Operations

• describe_default_cache - Describe the currently configured default cache (typically DuckDB)
• list_cached_streams - List all streams available in the default cache
• sync_source_to_cache - Run a sync from a source connector to the default DuckDB cache
• run_sql_query - Run SQL queries against the default cache

Call for Feedback

We'd love to hear your input below:

1. Are you using MCP servers today?
2. Are you integrating your data and AI workloads?
3. Have you tried the PyAirbyte MCP? What was your experience?