HubSpot connector

The HubSpot connector is a marketing connector. It reads CMS site pages from HubSpot over the CMS API, delivers their title and meta description to Bowrain for translation, and writes approved text back to the same pages.

note

This connector is added in a project's Connectors view in the web app, or through the workspace connectors API described below. Saved credentials are write-only — used for sync, never displayed again. All connector operations require the manage connectors permission. It is one route into a workspace among several — see Connectors for the full row.

What it syncs

The connector works against HubSpot CMS site pages through the cms/v3/pages/site-pages API.

Direction	Content
Fetch (in)	For each site page: the HTML title, and the meta description (when the page has one).
Publish (out)	The HTML title and meta description of a previously fetched page.

The connector reads the page title and meta description only. Page body content, modules, blog posts, emails, and landing pages are not included.

Prerequisites and credentials

You need a HubSpot account with the CMS pages API available, and a private app access token that can read and write CMS pages.

The connector accepts the following configuration keys:

Key	Required	Description
api_key	Yes	A HubSpot private app access token. It is sent as a bearer token in the Authorization header.
name	No	A human-readable name for the connector instance.
id	No	A stable identifier for the connector instance. When omitted, it defaults to hubspot.

Setup

Add the connector from a project's Connectors view in the web app, or by posting its type and configuration to POST /api/v1/{workspace}/connectors. Connectors are scoped to the workspace they are added in.

{
  "type": "hubspot",
  "config": {
    "name": "Marketing pages",
    "api_key": "pat-xx-xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
  }
}

The response returns the connector's id, name, and category. Use that id for subsequent fetch, publish, status, and remove calls.

How sync works

Sync is explicit: content moves only when you fetch or publish — Fetch now and Publish in the Connectors view, or the endpoints below. The connector does not poll HubSpot on a schedule and does not receive HubSpot webhooks.

Fetch requests site pages and stores each page's title and meta description as blocks on the target project's main content stream. Call POST /api/v1/{workspace}/connectors/{id}/fetch with the connector and project:

{ "connector_id": "{id}", "project_id": "{project}" }

The response reports how many items were fetched. Fetching stores source content; it does not itself start translation. Translate the fetched content the way you would any project content — for example by running a flow or by letting the project's server.converge policy produce and review targets.

Publish sends the project's stored title and meta-description text back to the matching pages. Call POST /api/v1/{workspace}/connectors/{id}/publish with the same fields:

{ "connector_id": "{id}", "project_id": "{project}" }

Each page is updated in place with a PATCH to the CMS pages API, matched by its HubSpot page ID. The connector writes to the page object; it does not perform a separate HubSpot publish step, so changes may need to be published from within HubSpot to appear on the live site.

Call GET /api/v1/{workspace}/connectors/{id}/status to see the connector's current view of HubSpot.

Limitations

• Only the HTML title and meta description are synced. Page body content and modules are out of scope.
• Fetch reads the first page of site pages returned by the API and does not paginate; accounts with more pages are not fully covered.
• Publishing updates the page object through a PATCH and does not push the page live; a separate publish action in HubSpot may be required.