Skip to content

Reveal MCP

Find the table behind a CMBS loan history and inspect what one row means before querying it.

Which dataset contains the records I need, and what does one row mean?

People and agents deciding which dataset can answer a question.

Interface
MCP server
Access
Request access
Availability
Access by arrangement
Runs on
Hosted service

API key required

Request Reveal MCP access
Cairn roles
AI connectors, Public records library
Evidence roles
Source
Semantic context
Source meaning, Relationships
Contribution
Discover catalog datasets, tables and lineage before selecting records for a question.
Limit
Catalog metadata does not establish that two source concepts are equivalent or that a dataset answers every question.

Find a table

What it does

  • Find cataloged datasets, tables, columns, providers and files.
  • Look up a known table by name and inspect its physical identity and description.
  • Follow recorded relationships and lineage before choosing data.

Live catalog lookup, November 3, 2026

Look up PROPERTY_MART.LOAN_PERIOD_FACT before querying a CMBS loan's history.

The lookup found the table and described its grain as one loan observation per CIK, asset number and reporting-period end date. That keeps a loan's history distinct from a count of loans.

cmdrvl.com.metadata.lookup({"reference":"PROPERTY_MART.LOAN_PERIOD_FACT","limit":3})

This checks a registered table's identity and description. It does not read its balances or establish that every relevant dataset is cataloged.

Setup

Tell us your question, intended MCP client and data scope. We will confirm access through Loom or an agreed connection and supply credentials and client configuration. There is no self-serve signup.

Request Reveal MCP access

Coverage and freshness

The catalog records available to your provisioned account. Catalog entries describe data; their presence does not grant access to the underlying rows or establish source freshness.

Limits

  • Catalog counts measure registered resources, not real-world entities or data rows.
  • Inspect applied filters and returned caveats. An unresolved search term may leave a broader result set than intended.
  • Names, descriptions and recorded lineage can be incomplete. Confirm the live schema with Data MCP before querying.

Related tools and guides

Contact support