Matia MCP
Matia runs a remote MCP (Model Context Protocol) server that lets an AI assistant work with your Matia workspace directly.
Once connected, an assistant can look up an asset, trace lineage, check why a sync failed, read a monitor's history, or estimate the downstream impact of a schema change, all in a single prompt.
What you can do
Like the rest of Matia, tools are grouped by product pillar:
- Catalog & Lineage: search the catalog, resolve an asset, walk lineage, and list tags and schema changes.
- Integrations (ETL): read integrations, connections, configured streams, and sync runs, and trigger, pause, or resume an integration.
- Observability: read monitors, issues and their evidence, and a one-call workspace health summary. Enable or disable a monitor, and preview the downstream impact of a change before you make it.
- Reverse ETL: inspect the record outcomes of a reverse-ETL run: what succeeded, what was rejected, and what was invalid.
- Governance & Audit: read who changed what and when, what changed in the workspace, and update tags, descriptions, or issue status.
See Tool Reference for the full list.
How it works
Your AI assistant talks to the Matia MCP server, which talks to your Matia workspace on your behalf, under your existing Matia permissions:
There's no separate identity or permission set for MCP. It's the same account and the same roles you already use to sign in to Matia.
Using Matia's MCP alongside other MCP servers
Matia's MCP server works well alongside other MCP servers and tools your assistant has access to.
Combining them gives your assistant more of the context it needs to answer a question fully: pairing a Matia lineage lookup with a ticketing tool to open a follow-up task, or combining a workspace health check with your team's on-call calendar, for example.