Menu
Mattering Automations connector

A HubSpot MCP built on the API surface HubSpot will still support

Connect one private app token and 50 HubSpot nodes become workflow steps for visual automations and AI agents: contacts, companies, deals, and tickets with full search, plus associations, pipelines, owners, properties, engagements, and lists. Every route was confirmed against HubSpot's current date-based API, the surface HubSpot is carrying forward as its legacy versions retire.

50 ready-made nodes One credential to connect Works with Claude, ChatGPT, and Matter

What the HubSpot MCP connector does

The HubSpot connector for Mattering Automations is 50 nodes across the CRM: contacts, companies, deals, and tickets each with list, get, create, update, search, and archive, plus associations with label discovery, pipelines and their stages, owners, property definitions, engagements (notes, tasks, logged calls, and logged emails), and lists with membership reads and manual-list membership writes.

Every node speaks MCP, the open protocol for connecting AI assistants to tools, so Claude, ChatGPT, and any MCP-capable assistant can use them, alongside Matter, the AI assistant built into the platform. An assistant can answer which deals moved stage this week, pull the contacts on an account, or log a note on a record's timeline, without anyone opening HubSpot.

Because Mattering is a work platform rather than middleware, CRM work lands somewhere useful by default: a scheduled deal-stage sweep can open prep tasks in Mattering Productivity when deals advance, and new HubSpot contacts can sync into a Mattering database for reporting next to the rest of the work. Mattering Sales is the platform's native CRM; this connector is for teams already living in HubSpot who want that data on the same canvas.

Two honesty notes are built into the node descriptions rather than buried in fine print. First, HubSpot deletes are archives: the four archive nodes move records to HubSpot's recycling bin, where they are restorable in the HubSpot UI for 90 days, and there is no restore API, so automations should treat them as deletes. Second, private app tokens carry granular scopes chosen at creation, so a connected account can still refuse individual nodes; HubSpot: Get Token Scopes shows exactly what a token can do.

One connection, three surfaces

  • Visual automations. Drag HubSpot nodes onto the canvas and wire them to triggers, conditions, and other apps.
  • AI agents, yours or ours. These nodes speak MCP, the open protocol, so Claude, ChatGPT, and any MCP-capable assistant can use them, and Matter, the agent built into the platform, uses them natively.
  • The whole platform. Combine with native tasks, projects, and other connectors, no middleware in between.

Every HubSpot node, ready to drop on the canvas

50 nodes, generated from the connector itself so this list always matches what ships. Expand a category to browse.

Read and query nodes 29

HubSpot: Get Account Details

Return the HubSpot account the token belongs to: portalId (hub id), account type, time zone, company currency, UI domain, and data hosting location. Needs no configuration and no CRM scopes, which makes it the connector's credential health check: a bad token fails here with a structured 401 before anything else runs.

HubSpot: Get Company

Fetch one company by record id, or by a unique property value when idProperty is set (for example domain). Name the properties you need; the default set is small.

HubSpot: Get Contact

Fetch one contact by record id, or by a unique property value when idProperty is set (for example email). Name the properties you need; the default set is small.

HubSpot: Get Deal

Fetch one deal by record id, or by a unique property value when idProperty is set (for example a custom unique deal property). Name the properties you need; the default set is small.

HubSpot: Get List

Fetch one list by its list id: name, object type, processing type (MANUAL, DYNAMIC, SNAPSHOT), size, and optionally its filter definition.

HubSpot: Get Owner

Fetch one owner by owner id, with email, name, userId, and team membership.

HubSpot: Get Property

Fetch one property definition by internal name: type, field type, enumeration options, and whether it enforces unique values.

HubSpot: Get Ticket

Fetch one ticket by record id, or by a unique property value when idProperty is set (for example a custom unique ticket property). Name the properties you need; the default set is small.

HubSpot: Get Token Scopes

Return the private app token's metadata: hub id, app id, the acting user id, and the token's ACTUAL granted scope list. Private app scopes are chosen when the app is created, so a connected HubSpot account can still return 403 MISSING_SCOPES on individual nodes; run this node to see exactly which scopes the token carries. The response contains no credential material.

HubSpot: List Association Labels

List the association types defined between two object types (HubSpot-defined and custom labels), each with its typeId, category, and label. This is the schema-discovery node for the associationTypeId values used in record-create associations arrays.

HubSpot: List Associations

List the records of one type associated to a given record: each result carries the associated record id plus the association's category, type id, and label. Example: all deals associated to contact 512.

HubSpot: List Calls

List call engagements (object type id 0-48), cursor paginated. Ask for hs_call_title, hs_call_status, and hs_call_duration via properties.

HubSpot: List Companies

List company pages (object type id 0-2), cursor paginated via paging.next.after. Returns the default property set unless you name properties explicitly. For filtered reads use HubSpot: Search Companies instead. Requires the crm.objects.companies.read scope on the private app token.

HubSpot: List Contacts

List contact pages (object type id 0-1), cursor paginated via paging.next.after. Returns the default property set unless you name properties explicitly. For filtered reads use HubSpot: Search Contacts instead. Requires the crm.objects.contacts.read scope on the private app token.

HubSpot: List Deals

List deal pages (object type id 0-3), cursor paginated via paging.next.after. Returns the default property set unless you name properties explicitly. For filtered reads use HubSpot: Search Deals instead. Requires the crm.objects.deals.read scope on the private app token.

HubSpot: List Logged Emails

List logged email engagements (object type id 0-49), cursor paginated. Ask for hs_email_subject and hs_email_status via properties.

HubSpot: List Memberships

List the record ids that are members of a list, cursor paginated, with each record's membership timestamp. Feed the ids into Get Contact or a batch step downstream.

HubSpot: List Notes

List note engagements (object type id 0-46), cursor paginated. Ask for hs_note_body explicitly via properties; add associations=contacts to see which records each note sits on.

HubSpot: List Owners

List the account's owners (users records can be assigned to): id, email, name, and teams. Use the id value (not userId) in hubspot_owner_id properties. Requires the crm.objects.owners.read scope.

HubSpot: List Pipeline Stages

List the stages of one pipeline in display order, with each stage's id, label, and metadata (deal win probability or ticket open/closed state). Use the stage ids when creating or moving deals and tickets.

HubSpot: List Pipelines

List every pipeline for an object type (deals or tickets, plus leads and some commerce objects on higher tiers), each with its stages inline. The id values here are what deal and ticket records store in their pipeline and stage properties.

HubSpot: List Properties

List every property defined on an object type, standard and custom: internal name, label, type, field type, group, and the options list for enumerations. This is the schema-discovery node: run it once to learn the exact property names the create, update, and search nodes expect.

HubSpot: List Tasks

List task engagements (object type id 0-27), cursor paginated. Ask for hs_task_subject, hs_task_status, and hs_timestamp via properties to build due-date automations.

HubSpot: List Tickets

List ticket pages (object type id 0-5), cursor paginated via paging.next.after. Returns the default property set unless you name properties explicitly. For filtered reads use HubSpot: Search Tickets instead. Requires the crm.objects.tickets.read scope on the private app token.

HubSpot: Search Companies

Search companies with property filters and free text. Searchable default text properties include name, domain, and phone. Filter groups OR together; filters inside a group AND together. Limits: 200 results per page, 10,000 results per query, and 5 search requests per second per account (shared across all search nodes). Archived records never appear; fresh writes can take a moment to be searchable.

HubSpot: Search Contacts

Search contacts with property filters and free text. Searchable default text properties include email, firstname, lastname, phone, and company. Filter groups OR together; filters inside a group AND together. Limits: 200 results per page, 10,000 results per query, and 5 search requests per second per account (shared across all search nodes). Archived records never appear; fresh writes can take a moment to be searchable.

HubSpot: Search Deals

Search deals with property filters and free text. Searchable default text properties include dealname. Filter groups OR together; filters inside a group AND together. Limits: 200 results per page, 10,000 results per query, and 5 search requests per second per account (shared across all search nodes). Archived records never appear; fresh writes can take a moment to be searchable.

HubSpot: Search Lists

Search the account's lists by name. Filter by processing type to find the MANUAL lists that the membership add and remove nodes can write to (DYNAMIC lists are filter-driven and cannot be edited via the API).

HubSpot: Search Tickets

Search tickets with property filters and free text. Searchable default text properties include subject and content. Filter groups OR together; filters inside a group AND together. Limits: 200 results per page, 10,000 results per query, and 5 search requests per second per account (shared across all search nodes). Archived records never appear; fresh writes can take a moment to be searchable.

Action nodes 21

HubSpot: Add List Members

Add records to a MANUAL or SNAPSHOT list by record id. DYNAMIC lists are filter-driven and reject membership writes; edit their filters in HubSpot instead. Reversible via HubSpot: Remove List Members.

HubSpot: Archive Company

DESTRUCTIVE: archive (delete) a company. HubSpot moves the record to the recycling bin, where it can be restored in the HubSpot UI for 90 days; after that it is permanently deleted. There is no restore API, so treat this as a delete in automations.

HubSpot: Archive Contact

DESTRUCTIVE: archive (delete) a contact. HubSpot moves the record to the recycling bin, where it can be restored in the HubSpot UI for 90 days; after that it is permanently deleted. There is no restore API, so treat this as a delete in automations.

HubSpot: Archive Deal

DESTRUCTIVE: archive (delete) a deal. HubSpot moves the record to the recycling bin, where it can be restored in the HubSpot UI for 90 days; after that it is permanently deleted. There is no restore API, so treat this as a delete in automations.

HubSpot: Archive Ticket

DESTRUCTIVE: archive (delete) a ticket. HubSpot moves the record to the recycling bin, where it can be restored in the HubSpot UI for 90 days; after that it is permanently deleted. There is no restore API, so treat this as a delete in automations.

HubSpot: Create Association

Associate two existing records with the default (unlabeled) association type for the pair, for example link contact 512 to company 1024. For labeled associations use the association type ids from HubSpot: List Association Labels via the extra-body pattern on record create/update instead.

HubSpot: Create Company

Create a company. Properties go in one JSON object keyed by internal property name; associate it to existing records in the same call via the associations array. Requires the crm.objects.companies.write scope.

HubSpot: Create Contact

Create a contact. Properties go in one JSON object keyed by internal property name; associate it to existing records in the same call via the associations array. Requires the crm.objects.contacts.write scope.

HubSpot: Create Deal

Create a deal. Properties go in one JSON object keyed by internal property name; associate it to existing records in the same call via the associations array. Requires the crm.objects.deals.write scope.

HubSpot: Create Note

Log a note on the CRM timeline and pin it to records via the associations array (a note without associations appears on no record). hs_timestamp is required and places the note on the timeline.

HubSpot: Create Task

Create a CRM task. hs_timestamp is required and is the task's DUE date; assign it with hubspot_owner_id and attach it to records via the associations array.

HubSpot: Create Ticket

Create a ticket. Properties go in one JSON object keyed by internal property name; associate it to existing records in the same call via the associations array. Requires the crm.objects.tickets.write scope.

HubSpot: Log Call

Log a call engagement on the CRM timeline (records that a call happened; does not place calls). hs_timestamp is required; attach the call to records via the associations array.

HubSpot: Log Email

Log an email engagement on the CRM timeline. This RECORDS an email that was sent or received elsewhere; it does NOT send email. hs_timestamp is required; attach it to records via the associations array.

HubSpot: Remove Association

Remove ALL associations between two records (both records themselves are untouched). Reversible: re-creating the association restores the link, though custom labels must be re-applied.

HubSpot: Remove List Members

Remove records from a MANUAL or SNAPSHOT list by record id. The records themselves are untouched; only the list membership changes, and re-adding restores it exactly.

HubSpot: Update Company

Update a company by record id (or by a unique property via idProperty). Partial update: only the properties you send change; clear a property by sending an empty string for it.

HubSpot: Update Contact

Update a contact by record id (or by a unique property via idProperty). Partial update: only the properties you send change; clear a property by sending an empty string for it.

HubSpot: Update Deal

Update a deal by record id (or by a unique property via idProperty). Partial update: only the properties you send change; clear a property by sending an empty string for it.

HubSpot: Update Task

Update a task: complete it ({"hs_task_status": "COMPLETED"}), reassign it, change its due date, or edit its text. Partial update; only named properties change.

HubSpot: Update Ticket

Update a ticket by record id (or by a unique property via idProperty). Partial update: only the properties you send change; clear a property by sending an empty string for it.

Connecting HubSpot takes one credential

HubSpot issues private app access tokens from Settings > Integrations > Private Apps in your HubSpot account. Create a private app, pick its scopes by checkbox, and the token appears under the app's Auth tab; it is sent as a bearer credential on every request. Scopes are granular (crm.objects.contacts.read, crm.objects.deals.write, and so on) and can be added to the app later without changing the token.

1

In HubSpot, go to Settings > Integrations > Private Apps, create an app, and check the scopes your automations need: each core object family has a read and a write scope, owner reads need crm.objects.owners.read, and list nodes need crm.lists.read or crm.lists.write.

2

Copy the access token from the app's Auth tab.

3

In Mattering Automations, open Connectors, click Add Connector, choose HubSpot, and create a credential with that token.

4

Click Test Connection. Mattering calls HubSpot's account details endpoint, which needs no CRM scopes, and all 50 HubSpot nodes appear in the canvas palette. Run HubSpot: Get Token Scopes next to confirm the token carries the scopes your workflows will actually use.

Common uses for the HubSpot MCP

Schema discovery an agent runs first

HubSpot: List Properties returns every property on an object type, standard and custom, with internal names, types, and enumeration options: exactly what the create, update, and search nodes expect. Add HubSpot: List Pipelines and HubSpot: List Pipeline Stages for stage ids, HubSpot: List Owners for assignment ids, and HubSpot: List Association Labels for association type ids, and an agent knows the shape of your CRM before it writes anything.

Search with HubSpot's full filter grammar

HubSpot: Search Contacts, HubSpot: Search Companies, HubSpot: Search Deals, and HubSpot: Search Tickets carry the complete filter grammar: filter groups OR together while filters inside a group AND together, with a full operator set and free-text search over the default text properties. This is also the polling pattern: run a search on a schedule with a last-modified date filter and act on what changed.

Deal-stage sweeps that open prep tasks

HubSpot: Search Deals finds the deals that entered a stage, HubSpot: List Associations pulls the contacts on each one, and HubSpot: Get Contact fetches their details. Route the results into Mattering Productivity as prep tasks and the pipeline stops depending on someone remembering to check the board.

Contacts synced into a Mattering database

A scheduled HubSpot: Search Contacts with a last-modified date filter feeds new and changed contacts into a Mattering database, so reporting, enrichment, and cross-app automations work from a copy that lives next to the rest of the work, without exporting CSVs.

Timeline logging without the copy-paste

HubSpot: Create Note, HubSpot: Create Task, HubSpot: Log Call, and HubSpot: Log Email write engagements onto record timelines via the associations array. The log-email node records an email that was sent or received elsewhere; it never sends one. Workflows elsewhere on the canvas can keep the CRM's history complete as a side effect.

List membership as a workflow step

HubSpot: Search Lists filters lists by processing type, and HubSpot: Add List Members and HubSpot: Remove List Members write membership on MANUAL and SNAPSHOT lists by record id. DYNAMIC lists are filter-driven and reject membership writes, so the search node's processing-type filter is how a workflow finds the lists it can actually edit.

Example: Deals that change stage open prep tasks, with the research attached

Built on the Automations canvas in a few minutes, or described in plain language to Matter and built for you.

1

Trigger: a schedule runs the sweep each morning.

2

HubSpot: Search Deals filters for deals in the target stage whose last-modified date falls inside the polling window, with the stage and date filters ANDed together in one filter group.

3

HubSpot: List Associations returns the contact ids associated to each matched deal.

4

HubSpot: Get Contact fetches each contact's name, email, and title, named explicitly in properties since the default property set is small.

5

A task opens in Mattering Productivity with the deal name, amount, and contact roster, assigned to whoever preps the call.

6

HubSpot: Create Note logs a note back onto the deal's timeline via the associations array, so the CRM shows the prep happened.

HubSpot MCP questions, answered

What HubSpot operations are supported?

50 nodes: contacts, companies, deals, and tickets each with list, get, create, update, search, and archive; associations (list, create, remove, plus label discovery); pipelines and stages; owners; property definitions; engagements (create and list notes, tasks, calls, and logged emails, plus task updates); and lists (search, get, memberships, plus membership add and remove on manual lists). Deliberately not shipped: batch endpoints, custom object CRUD, property and pipeline and list administration, GDPR delete (permanent, bypasses the recycling bin), marketing surfaces such as email send and forms and workflows, and webhook subscriptions.

Does this work with Claude, ChatGPT, or other AI assistants?

Yes. Mattering exposes these nodes over MCP, an open protocol for connecting AI assistants to tools, so any MCP-capable assistant, including Claude and ChatGPT, can connect to your workspace and use them. Matter, the assistant built into the platform, uses the same nodes natively.

Why do some nodes return 403 when Test Connection passed?

Because private app tokens carry granular scopes chosen when the app is created, and the connection test uses an endpoint that needs no CRM scopes at all. A token missing crm.objects.deals.read connects cleanly and then returns 403 with category MISSING_SCOPES on deal reads. Run HubSpot: Get Token Scopes to see the token's actual granted scope list, then add the missing scopes in the app's settings in HubSpot; the token string does not change, so no reconnect is needed.

Does archiving a record delete it?

Almost. HubSpot's record delete is an archive: the four archive nodes move contacts, companies, deals, and tickets to HubSpot's recycling bin, where they can be restored in the HubSpot UI for 90 days, after which they are permanently deleted. There is no restore API, so an automation cannot undo an archive; the nodes are flagged destructive so they can be gated behind a condition or approval.

What are the search limits?

HubSpot's search API returns up to 200 results per page and 10,000 results per query, and allows 5 search requests per second per account, shared across all four search nodes. The grammar allows up to 5 filter groups of 6 filters each, 18 filters total, and one sort rule. Archived records never appear in search, and freshly written records take a short indexing delay to become searchable. For sweeps that could exceed 10,000 results, narrow the window with a last-modified date filter.

Can this connector send email?

No. HubSpot: Log Email records an email engagement on a record's timeline, capturing a message that was sent or received elsewhere; it never causes an email to be delivered. HubSpot's marketing email send surfaces are a different product family and were deliberately left out of this CRM connector.

Can a HubSpot change trigger a workflow?

Not as an inbound trigger node. Webhook subscriptions are app-level configuration owned by the private app UI, so they were deliberately left out. The honest pattern today is polling: run one of the search nodes on a schedule with a last-modified date filter and act on what changed since the previous run.

Which HubSpot API version does this use?

The current date-based surface: every CRM route uses HubSpot's 2026-03 release, the versioning model HubSpot introduced to replace its legacy v3 and v4 paths, which stay supported only until their announced sunset. Each route in this connector was confirmed against the current surface at build time, so workflows built here are on the API HubSpot will still support and will not need a migration when the legacy paths retire.

50 nodes waiting on the canvas

Put HubSpot to work inside your workspace

Connect once, then use every node in visual automations and AI agent workflows across the Mattering platform.

All product names, logos, and brands are property of their respective owners and are used for identification purposes only. Use of these names, logos, and brands does not imply endorsement or affiliation. HubSpot is a product of its respective owner; Mattering is an independent platform.

Start Free Explore our Products