> ## Documentation Index
> Fetch the complete documentation index at: https://honeydew.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Collibra

[Collibra](https://www.collibra.com/) is a data governance platform. It catalogs the assets
across your data landscape — along with their owners, policies, and stewardship workflows — and
governs how they are used. The integration publishes Honeydew's semantic definitions into
Collibra as governed assets.

<Info>
  This integration is in Beta. To enable for your organization,
  please contact [support@honeydew.ai](mailto:support@honeydew.ai).
</Info>

## Two Context Layers

Honeydew and Collibra sit at different levels of the stack:

| | Honeydew [Context Layer](/docs/integration/context-layer/overview) | Collibra |
| - | - | - |
| **Grounded in** | The semantic layer — governed metrics, entities, and calculations | The catalog — assets, policies, ownership, and stewardship |
| **Scope** | How your organization analyzes: methodologies, conventions, and past decisions | What exists across the data landscape, who owns it, and the rules that apply |
| **Answers** | How a number is defined and computed correctly | What an asset is, who is accountable for it, and whether it is approved |

The two are complementary. Collibra governs the assets and the policies around them; Honeydew
defines the logic that turns them into governed metrics, and generates the SQL. Publishing
Honeydew definitions into Collibra puts them under the same stewardship as the rest of the
catalog.

## What Gets Published

Each workspace branch becomes one data model in Collibra, holding the entities and fields it
defines. Honeydew publishes into Collibra's out-of-the-box asset types, so there are no custom
asset types to install:

| Honeydew | Collibra asset type |
| - | - |
| Workspace branch | `Data Model` |
| Entity | `Data Entity` |
| Attribute | `Data Attribute` |
| Metric | `Metric` |

Every asset carries its display name and its business [metadata](/docs/governance/metadata) —
description, owner, labels, and folder. Attributes and metrics also carry their SQL expression,
data type, and any synonyms and sample values curated for the AI.

## Setup

Register a Collibra application, then configure the integration in Honeydew.

### Register a Collibra application

Collibra authenticates integrations with OAuth client credentials. In Collibra:

1. Go to **Settings** → **OAuth applications**.
2. Click **+ New Application**.
3. In **Select your application type**, click **Register integration**. Integration
   applications use a machine-to-machine connection and need no user login.
4. Enter an **Application name**, such as `honeydew`, and click **Register**.
5. Copy the **Client ID** and **Client Secret**.

<Warning>
  The client secret is shown only at registration and cannot be retrieved afterwards. If it is
  lost, register a new application.
</Warning>

Collibra attributes the integration's changes to an application-specific system user, so
published assets are traceable to the integration rather than to a person.

See Collibra's
[OAuth Applications documentation](https://productresources.collibra.com/docs/collibra/latest/Content/Settings/OAuth/co_oauth-settings.htm)
for application management and secret rotation.

### Configure Honeydew

1. Go to **Settings** → **Semantic integrations**.
2. Under **Add new integration**, select **Collibra**.
3. Enter a **Connection identifier** — a name for this credential set, for example
   `collibra_prod`.
4. Enter the **Instance URL** of your Collibra instance, in the form
   `https://<instance>.collibra.com`.
5. Paste the **Client ID** and **Client secret** and save.

## Publish a Workspace

Publishing requires the **Editor** [role](/docs/access-control/user-access-control#user-roles).

1. Click the branch, and in the menu click **Branch integrations** → **Collibra** →
   **Publish to Collibra...**
2. Select the **Community** and **Domain** to publish into.
3. Select the **Warehouse database** — the Collibra database the source tables were registered
   under, which is what lineage is drawn against. Leave it empty to publish without lineage.
4. Set the **Model name**, or select an existing data model to republish into it. The default is
   the workspace name, with the branch appended.
5. Click **Publish**, then open the published model from the link in the result.

The result reports how many assets were pushed, archived, and linked. It also reports source
columns that no registered column matched: those attributes are published without lineage, so a
count above zero usually means the warehouse database is the wrong one, or the tables have not
been registered in Collibra yet.

## Where Lineage Appears

Honeydew links each published `Data Attribute` to the `Column` it reads, using Collibra's
**represents / represented by** relation. Open an attribute in Collibra and select **Diagram**
to see it with the columns it represents. A calculated attribute resolves through the attributes
it depends on, down to those columns.

## Republishing

Publishing again updates the data model in place. Assets are keyed on their full name within the
target domain, so republishing an unchanged workspace changes nothing.

A field deleted in Honeydew is archived in Collibra on the next publish, and stays recoverable.
Changing the model name publishes a second data model rather than renaming the first, since the
name is part of every asset's identity. Publishing into a different domain does the same.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.