Skip to main content
DataGalaxy The DataGalaxy app exposes the DataGalaxy data catalog API. It can be consumed two ways: as a remote MCP server that Agent Factory agents call as tools, or as a Builder app whose instructions you call directly from DSUL. It covers the full catalog — dictionary (sources, containers, structures, fields), glossary, usages, data processings — plus links, comments and tasks, and a global search across every object. Authentication is a single DataGalaxy Personal Access Token (PAT) or integration token stored in the workspace configuration: the connector exchanges it server-side for a short-lived access token. The credential is never exposed to the agent — agents are identified by the capability Scope context_id,agent_id,user_id and the connector resolves the token from its own app configuration.

Catalog dictionary

Create, read, update and delete sources, containers, structures and fields, with the generic object operations for any entity type.

Glossary, usages & data processing

Manage glossary properties (business terms, concepts, indicators), usages (applications, dashboards, datasets) and data flows / processings.

Links & collaboration

Relate any two catalog entities, post comments and manage tasks across the catalog, with full-text search over everything.

Who is this for?

This connector is used by three different roles. Jump to the section that matches yours — each one is self-contained.

Agent builder

You build agents in Agent Factory and want them to read and write the DataGalaxy catalog. → Agent builder tab.

Platform admin

You run the platform and want to publish DataGalaxy as a reusable capability. → Platform admin setup accordion below.

Workspace builder

You write Builder automations (DSUL) that call DataGalaxy operations directly. → Workspace builder tab.

Prerequisites

  • A DataGalaxy account with API access to the workspaces (client spaces) you intend to manage.
  • A Personal Access Token (PAT) or integration token — generated in DataGalaxy under your account / integration settings. The connector exchanges it server-side for a short-lived access token.
  • Your DataGalaxy API base URL — default https://api.datagalaxy.com (override only for a region-specific or self-hosted instance).
  • The token’s user must hold the DataGalaxy roles needed for the operations you call (read for list/get, write for create/update/delete on the relevant workspace version).
Goal: DataGalaxy is a per-workspace connector — each workspace pastes its own DataGalaxy token in the app configuration (see the Workspace builder tab), so there is no platform-wide credential to provision and no central OAuth client.
There is no shared DataGalaxy credential. The PAT / integration token always lives in the consuming workspace’s app configuration and is resolved server-side. A Governance capability you publish here points at a specific workspace’s MCP endpoint; that workspace still owns the credential.

Declare the capability in AI Governance (optional)

1

Open AI Governance > Capabilities

Create (or edit) the DataGalaxy capability.
2

Point it at the MCP endpoint

Set the capability’s MCP server URL to the connector’s MCP Endpoint (the workspace running the connector), and set its Scope to:
context_id,agent_id,user_id
The agent_id in the scope is what lets the connector identify the calling agent.
3

Make it available to agent builders

Once created, the capability appears in the capability picker for agent builders in your organization, who enable it on their agents. Access to the catalog follows your organization’s existing roles; there is no per-capability role grant for this connector.
Declaring the capability makes the connector available; it does not provision any credential. This connector resolves a single DataGalaxy token from the consuming workspace’s app configuration — there is no OAuth auth-config JSON to attach in Governance and no per-user OAuth flow.

Agent builder

Goal: let an agent you build in Agent Factory read and write the DataGalaxy catalog through MCP tools.
Before an agent can call the connector, a Workspace builder must have installed and configured the DataGalaxy app in a workspace (see the Workspace builder tab). Optionally, a Platform admin may have published a DataGalaxy capability in AI Governance (see the Platform admin setup accordion above).
The DataGalaxy credential is resolved server-side from the app configuration — never exposed to the agent. Your agent is identified by the agent_id that Agent Factory injects through the capability Scope.
1

Install and configure the connector in a workspace

Follow the Workspace builder tab: install DataGalaxy in a workspace and provide the PAT / integration token and (optionally) the base URL.
2

Add the MCP capability to your agent

In your agent, add a capability pointing at that workspace’s MCP Endpoint URL, and set its Scope to:
context_id,agent_id,user_id
The agent_id is what lets the connector identify your agent.
3

Brief the agent

Tell the agent the tools exist and when to reach for them. Copy-pasteable starter:
You have access to the DataGalaxy MCP server. Use it whenever the user asks about catalog content — sources, containers, structures (tables), fields, glossary terms, usages, data processings, links, comments or tasks. Each tool takes an `action` argument selecting the concrete operation. Examples: "Find all glossary terms tagged GDPR" (catalog search), "List the structures of the Snowflake source", "Add a comment to the customer table", "Create a task to review the orders dataset". DataGalaxy IDs use the DoubleUuid format and most operations need a `versionId` (use the catalog tool's listWorkspaces / listVersions actions to discover it). Prefer calling a tool over guessing, and confirm with the user before any destructive action (delete a structure, remove a glossary term).
Refine the trigger keywords (source names, glossary domains, project tags) so the agent reliably picks up the right intent in your context.
Legacy AI Knowledge agents (no native MCP picker): add the connector under Advanced > Tools > MCP and paste the MCP Endpoint URL. The DataGalaxy token is still resolved server-side from the installed app configuration.

Available Tools

Tools are entity-grouped: each tool takes an action argument selecting the concrete operation, plus the per-action parameters. Most operations need a versionId (a DataGalaxy workspace version) — discover it with the catalog tool.
ToolDescription
catalogCatalog metadata + global search. Actions: listWorkspaces, listVersions, listAttributes, search.
sourcesSources — top-level dictionary entities. Actions: list, get, create, update, delete.
containersContainers (models, directories). Actions: list, get, create, update, delete.
structuresStructures (tables, datasets, views). Actions: list, get, create, update, delete.
fieldsFields (columns of a structure). Actions: list, get, create, update, delete.
propertiesGlossary properties (business terms, concepts, indicators, …). Actions: list, get, create, update, delete.
usagesUsages (applications, dashboards, datasets, …). Actions: list, get, create, update, delete.
dataProcessingsData flows and processings. Actions: list, get, create, update, delete.
linksRelationships between entities. Actions: get, create, delete.
commentsComments on an entity (rich-text HTML). Actions: list, create.
tasksTasks assigned across the catalog. Actions: list, create, update.
objectsGeneric object operations for any entity type. Actions: get, update.

Output Formats

Every tool accepts an outputFormat argument that controls the MCP response shape:
  • verbose (default) — human-readable text optimized for LLM consumption
  • structured — concise machine-readable JSON in structuredContent
  • both — the structured payload, with its JSON also rendered as text

Tool Details

Global full-text search across every catalog object.
{
  "name": "catalog",
  "arguments": {
    "action": "search",
    "query": "customer",
    "versionId": "abcd-1234...",
    "limit": 50,
    "filters": { "Status": ["Validated"] }
  }
}
ParameterRequiredDescription
actionYesOne of listWorkspaces, listVersions, listAttributes, search.
queryFor searchSearch query text.
versionIdNoScope the search to a workspace version.
limitNoMax results.
filtersNoAttribute filter object.
workspaceIdFor listVersionsWorkspace (client space) id.

sources — create

{
  "name": "sources",
  "arguments": {
    "action": "create",
    "versionId": "abcd-1234...",
    "name": "Customer Database",
    "type": "Database",
    "technicalName": "customer_db",
    "owners": ["alice@example.com"],
    "status": "Validated"
  }
}
ParameterRequiredDescription
actionYesOne of list, get, create, update, delete.
versionIdYesWorkspace version id.
nameFor createSource display name.
typeFor createEntity type code (DataGalaxy-specific).
sourceIdFor get/update/deleteSource id (UUID).
technicalName / description / summary / status / owners / stewards / tagsNoStandard catalog attributes (status: Proposed, InRevision, Validated, InValidation, Obsolete).

fields — create

Creates a field (column) under a structure.
{
  "name": "fields",
  "arguments": {
    "action": "create",
    "versionId": "abcd-1234...",
    "structureId": "struct-id",
    "name": "email",
    "type": "Column",
    "columnDataType": "varchar",
    "size": 255
  }
}
ParameterRequiredDescription
actionYesOne of list, get, create, update, delete.
versionIdYesWorkspace version id.
structureIdFor createParent structure id.
name / typeFor createField name and type code (Column, Field).
columnDataTypeNoFree-form data type (e.g. varchar, int, date).
sizeNoColumn size.
fieldIdFor get/update/deleteField id (UUID).
{
  "name": "links",
  "arguments": {
    "action": "create",
    "versionId": "abcd-1234...",
    "fromId": "field-id",
    "toId": "business-term-id",
    "type": "implements"
  }
}
ParameterRequiredDescription
actionYesOne of get, create, delete.
versionIdYesWorkspace version id.
fromIdYesSource entity id.
toIdFor create/deleteTarget entity id.
typeFor createLink type code.
linkTypeFor deleteLink type code to remove.

comments — create

Rich-text HTML is supported in the content field.
{
  "name": "comments",
  "arguments": {
    "action": "create",
    "versionId": "abcd-1234...",
    "entityId": "object-id",
    "content": "<p>Reviewed and validated.</p>"
  }
}
ParameterRequiredDescription
actionYeslist or create.
versionIdYesWorkspace version id.
entityIdYesEntity the comment is attached to.
contentFor createComment body (rich-text HTML supported).

Error Handling

HTTP StatusErrorSolution
401UnauthorizedVerify the PAT / integration token and baseUrl. The token is exchanged server-side; an invalid token fails the exchange.
403ForbiddenCheck the DataGalaxy role and scopes on the workspace version.
404Not FoundVerify versionId, objectId and any parent IDs (DoubleUuid format).
422UnprocessableA required attribute is missing or an enum value (type, status) is invalid.
429Rate LimitedBack off and retry with exponential delay.
500Server ErrorTransient DataGalaxy error — retry shortly.

Common Issues

“DataGalaxy not configured” — No token could be resolved. Complete the app install with a valid token, call configureDataGalaxy to store one in the session, or pass an Authorization: Bearer header. “DataGalaxy credentials exchange failed” — The provided PAT / integration token was rejected when exchanging it for an access token. Re-generate the token in DataGalaxy and re-paste it in the app config; confirm the base URL points at the right region. “The calling agent could not be identified” — The MCP capability Scope does not declare agent_id. Set the Scope to context_id,agent_id,user_id on the capability. versionId errors — Most operations are scoped to a workspace version. Use the catalog tool’s listWorkspaces then listVersions actions (or the listWorkspaces / listVersions instructions) to obtain a valid versionId before calling entity operations.

External Resources

DataGalaxy API

Official DataGalaxy API documentation.

Tool Agents

Learn how Agent Factory agents consume MCP tools in Prisme.ai.