MCP server for the Airbyte Public API. Built with the official MCP Python SDK (FastMCP).
Lets any MCP-compatible client (Cursor, Claude Desktop, Claude Code, MCP Inspector, etc.) interact with your Airbyte instance through natural language.
- 36 tools covering workspaces, sources, destinations, connections, jobs, refresh/clear, job logs, tags, streams, and connector definitions
- Read and write operations for core resources (create, update, delete)
- Job diagnostics via the internal Configuration API (self-managed): detailed failure reasons, per-stream stats, and structured logs
- Cloud full-text sync logs via
airbyte_get_cloud_sync_logs(Airbyte Cloud only; parity with official Replication MCP) - Automatic token exchange with in-memory caching and transparent 401 retry
- Markdown and JSON response formats on summary tools; JSON-only for log tools
- Pagination support (limit/offset) on all list tools
- Runs over stdio (local) — works with Cursor, Claude Desktop, Claude Code, and Docker
- Works with self-managed Airbyte (abctl) and Airbyte Cloud
| Tool | Description |
|---|---|
| Health | |
airbyte_health_check |
Ping the Airbyte API |
| Workspaces | |
airbyte_list_workspaces |
List workspaces with pagination |
airbyte_get_workspace |
Get workspace details by ID |
| Sources | |
airbyte_list_sources |
List source connectors (filter by workspace) |
airbyte_get_source |
Get source details by ID |
airbyte_create_source |
Create a new source connector |
airbyte_update_source |
Update an existing source |
| Destinations | |
airbyte_list_destinations |
List destination connectors (filter by workspace) |
airbyte_get_destination |
Get destination details by ID |
airbyte_create_destination |
Create a new destination connector |
airbyte_update_destination |
Update an existing destination |
| Connections | |
airbyte_list_connections |
List connections / pipelines (filter by workspace) |
airbyte_get_connection |
Get connection details including stream config |
airbyte_create_connection |
Create a new connection (pipeline) |
airbyte_update_connection |
Update an existing connection |
| Jobs | |
airbyte_list_jobs |
List jobs (filter by connection, type, status, dates) |
airbyte_get_job |
Get job details (status, duration, bytes/rows synced) |
airbyte_trigger_sync |
Trigger a sync or reset job |
airbyte_cancel_job |
Cancel a running job |
airbyte_trigger_refresh |
Trigger a per-stream refresh (internal API, self-managed) |
airbyte_trigger_clear |
Clear destination data for streams (internal API, self-managed) |
airbyte_wait_for_job |
Poll until a job reaches a terminal status (internal API) |
| Job Logs (Internal API — self-managed only) | |
airbyte_list_jobs_internal |
List all job types including refresh and clear |
airbyte_get_job_details |
Per-attempt stats, failure reasons, and stacktraces |
airbyte_get_job_logs |
Structured log entries for all attempts |
airbyte_get_attempt_logs |
Structured log entries for a specific attempt |
| Job Logs (Cloud only — full text) | |
airbyte_get_cloud_sync_logs |
Full-text sync logs with pagination (Cloud Config API) |
| Streams | |
airbyte_get_stream_properties |
Get stream properties for a source/destination pair |
| Tags | |
airbyte_list_tags |
List tags |
airbyte_create_tag |
Create a tag |
airbyte_update_tag |
Update a tag |
airbyte_delete_tag |
Delete a tag |
| Connector Definitions | |
airbyte_list_source_definitions |
List source connector definitions |
airbyte_get_source_definition |
Get a source connector definition |
airbyte_list_destination_definitions |
List destination connector definitions |
airbyte_get_destination_definition |
Get a destination connector definition |
See docs/endpoints.md for the full Airbyte API endpoint checklist.
- A running Airbyte instance — either:
- Self-managed via abctl (see docs/local-setup.md)
- Airbyte Cloud
- One of the following to run the server:
- uvx (zero-install; runs the published package on demand — see Run with uvx), or
- uv + Python 3.13+ (local clone / development), or
- Docker (no Python / uv needed on the host)
git clone https://github.com/trustxai/airbyte-mcp.git
cd airbyte-mcp
uv synccp .env.example .envFor self-managed (abctl), retrieve credentials:
abctl local credentialsEdit .env with your client-id and client-secret. See docs/authentication.md for details.
uv run airbyte-mcpuvx runs the published PyPI package on demand — no
clone, no virtualenv, no persistent install. The command matches the package
name, so no --from is needed:
uvx airbyte-mcpNot the same as the official Airbyte replication MCP. That one lives in the
airbytepackage and is invoked asuvx --from=airbyte@latest airbyte-mcp. Here the PyPI package isairbyte-mcp, souvx airbyte-mcpunambiguously resolves to this server.
Credentials are passed via the client's env block (see below), or exported in
your shell for a manual run:
AIRBYTE_API_URL=http://localhost:8000/api/public/v1 \
AIRBYTE_CLIENT_ID=<your-client-id> \
AIRBYTE_CLIENT_SECRET=<your-client-secret> \
uvx airbyte-mcpPython version: the package targets Python 3.13+. uv auto-provisions a matching interpreter, so this normally just works. If your environment pins an older default, force it with
uvx --python=3.13 airbyte-mcp.
Client config (e.g. Cursor .cursor/mcp.json) using uvx:
{
"mcpServers": {
"airbyte": {
"command": "uvx",
"args": ["airbyte-mcp"],
"env": {
"AIRBYTE_API_URL": "http://localhost:8000/api/public/v1",
"AIRBYTE_CLIENT_ID": "<your-client-id>",
"AIRBYTE_CLIENT_SECRET": "<your-client-secret>"
}
}
}
}Every MCP client (Cursor, Claude Desktop, etc.) can run the server in one of three ways:
- uvx — zero-install; runs the published package on demand (see Run with uvx).
- uv — from a local clone; best for development.
- Docker — no Python / uv required on the host; everything runs in a container. Build the image once and every client config reuses it.
Host networking note: if Airbyte is running on your host machine (e.g. via
abctl), inside the Docker containerlocalhostdoes not point to your host. Usehttp://host.docker.internal:8000/api/public/v1on macOS/Windows, or add--network=hostto thedocker runargs on Linux.
docker build -t airbyte-mcp:latest .Add to .cursor/mcp.json (project-level) or ~/.cursor/mcp.json (global):
Option A — uv
{
"mcpServers": {
"airbyte": {
"command": "uv",
"args": ["--directory", "/path/to/airbyte-mcp", "run", "airbyte-mcp"],
"env": {
"AIRBYTE_API_URL": "http://localhost:8000/api/public/v1",
"AIRBYTE_CLIENT_ID": "<your-client-id>",
"AIRBYTE_CLIENT_SECRET": "<your-client-secret>"
}
}
}
}Option B — Docker
{
"mcpServers": {
"airbyte": {
"command": "docker",
"args": [
"run", "--rm", "-i",
"--name", "airbyte-mcp",
"-e", "AIRBYTE_API_URL",
"-e", "AIRBYTE_CLIENT_ID",
"-e", "AIRBYTE_CLIENT_SECRET",
"airbyte-mcp:latest"
],
"env": {
"AIRBYTE_API_URL": "http://host.docker.internal:8000/api/public/v1",
"AIRBYTE_CLIENT_ID": "<your-client-id>",
"AIRBYTE_CLIENT_SECRET": "<your-client-secret>"
}
}
}
}Add to claude_desktop_config.json:
Option A — uv
{
"mcpServers": {
"airbyte": {
"command": "uv",
"args": ["--directory", "/path/to/airbyte-mcp", "run", "airbyte-mcp"],
"env": {
"AIRBYTE_API_URL": "http://localhost:8000/api/public/v1",
"AIRBYTE_CLIENT_ID": "<your-client-id>",
"AIRBYTE_CLIENT_SECRET": "<your-client-secret>"
}
}
}
}Option B — Docker
{
"mcpServers": {
"airbyte": {
"command": "docker",
"args": [
"run", "--rm", "-i",
"--name", "airbyte-mcp",
"-e", "AIRBYTE_API_URL",
"-e", "AIRBYTE_CLIENT_ID",
"-e", "AIRBYTE_CLIENT_SECRET",
"airbyte-mcp:latest"
],
"env": {
"AIRBYTE_API_URL": "http://host.docker.internal:8000/api/public/v1",
"AIRBYTE_CLIENT_ID": "<your-client-id>",
"AIRBYTE_CLIENT_SECRET": "<your-client-secret>"
}
}
}
}claude mcp add \
--env AIRBYTE_API_URL=http://localhost:8000/api/public/v1 \
--env AIRBYTE_CLIENT_ID=<your-client-id> \
--env AIRBYTE_CLIENT_SECRET=<your-client-secret> \
--transport stdio \
airbyte \
-- uvx airbyte-mcpThe Inspector can launch the stdio server directly:
npx @modelcontextprotocol/inspector uvx airbyte-mcpSet AIRBYTE_API_URL, AIRBYTE_CLIENT_ID, and AIRBYTE_CLIENT_SECRET in the
Inspector's environment panel (or export them in your shell first).
If you just want to exercise the server from the CLI:
# uvx — no clone
uvx airbyte-mcp
# uv
uv run airbyte-mcp
# Docker
docker run --rm -i --env-file .env airbyte-mcp:latest| Variable | Required | Default | Description |
|---|---|---|---|
AIRBYTE_API_URL |
No | http://localhost:8000/api/public/v1 |
Airbyte API base URL |
AIRBYTE_CLIENT_ID |
Yes* | — | Application client ID |
AIRBYTE_CLIENT_SECRET |
Yes* | — | Application client secret |
AIRBYTE_ACCESS_TOKEN |
No | — | Pre-fetched token (skips exchange) |
AIRBYTE_INTERNAL_API_URL |
No | (derived) | Override Config API base (Cloud: https://cloud.airbyte.com/api/v1) |
AIRBYTE_REQUEST_TIMEOUT_SECONDS |
No | 30 |
Default HTTP timeout for public API calls |
AIRBYTE_INTERNAL_LOG_TIMEOUT_SECONDS |
No | 120 |
Timeout for internal/Cloud Config API log endpoints |
*Not required if AIRBYTE_ACCESS_TOKEN is provided.
- Authentication — token exchange, credentials setup
- Architecture — system design, package layout, token lifecycle
- Comparison vs Official MCPs — open-source-first positioning vs official MCPs
- Endpoints Checklist — full Airbyte API coverage status
- Local Setup — abctl installation walkthrough
- Contributing — development workflow, PR guidelines
- Security — vulnerability reporting
- Changelog — release history
Contributions are welcome! See docs/CONTRIBUTING.md to get started.
Apache-2.0 — see LICENSE for details.