CRMConnector
HubSpot
Search, create and update HubSpot contacts, companies, deals and tickets, and add notes, from Claude or ChatGPT, through Tulina's own MCP server.
Official site: app.hubspot.com
What it can do
2 read/3 write
- 01Read
Search and read contacts, companies, deals and support tickets, with pipeline stage names
- 02Write
Create, update, archive and merge records across the four objects
- 03Write
Add notes to a record
- 04Write
Create and manage lists, which HubSpot calls segments
- 05Read
Read properties, pipelines, owners and website visits
Four objects, contacts, companies, deals and tickets, all readable and writable, plus notes added to any of them. A workflow that needs to qualify a contact and log why does not need a second tool for that. The connector reads a portal's real property names and pipeline stages before it writes, so a create or update uses the values HubSpot accepts.
Does HubSpot have an MCP?
Yes. HubSpot hosts a remote MCP server at https://mcp.hubspot.com. Its developer documentation, last modified on 30 September 2026, labels it beta and says it needs OAuth with PKCE through an MCP connector that you create under Development in HubSpot. Through that server an assistant can read and write CRM records, and also work with activities, landing pages, blog posts, campaigns and marketing emails. Tulina reaches HubSpot through its own MCP server instead, on the same link that carries the rest of your stack.
Does HubSpot have API integration?
Yes. HubSpot exposes its CRM through APIs, and the Tulina connector calls them with a service key that you create. The key's scopes decide what the agent can touch, and Tulina's connection test reports which scope families the key holds.
How do you connect HubSpot to Claude and ChatGPT?
- In HubSpot, open Development, then Keys, then Service keys, and create a service key. Tick the CRM scopes you need: contacts, companies, deals and tickets each come as read and write, plus the schema scopes, lists and owners.
- Paste the key, which starts with
pat-, into Tulina's HubSpot connector. It is stored in the vault and never shown to the model. A private app token created before HubSpot ended private app creation pastes the same way. - Add
https://mcp.tulina.ai/mcpto Claude or ChatGPT and sign in with your Tulina account. The MCP page has the same link.
Several portals can be connected as named accounts, and a request names the one it is for.
Tulina vs HubSpot's official MCP server
Both reach your CRM. They differ in who holds the credential and how wide the surface is.
| HubSpot's MCP server | Tulina | |
|---|---|---|
| Endpoint | mcp.hubspot.com, hosted by HubSpot | mcp.tulina.ai/mcp, one link for every connected tool |
| Sign-in | OAuth with PKCE, through an MCP connector created in the account | A HubSpot service key held in Tulina's vault, for a person or the organisation |
| Permissions | Follow the signed-in user's own HubSpot permissions | Follow the scopes ticked on the key |
| CRM records | Contacts, companies, deals, tickets, leads and custom objects | Contacts, companies, deals, tickets and custom objects, plus lists, properties, pipelines and owners |
| Beyond the CRM | Activities, landing pages, blog posts, campaigns, marketing emails, conversations | Website visits |
| Duplicate merge | Not listed on the page | Merge with a dry run first |
| In the same conversation | HubSpot | HubSpot plus Apollo, Lemlist, Gmail, Slack and the other connected tools |
If the work lives inside HubSpot, such as publishing a landing page, drafting a marketing email or reading a conversation thread, HubSpot's own server covers more. Tulina earns its place when a deal update is one step among several tools.
What you can ask it
"Which deals over 20,000 are still open, and what stage is each in?" The agent reads the deal pipeline to learn which stages are open, then searches deals with an amount filter and asks for stage and pipeline labels next to the raw ids.
"Add Jane Doe at Acme as a contact and note that she asked for pricing." The agent lists the contact properties to get the internal names, finds the Acme company, creates the contact, associates the two and attaches the note to the new record.
"Merge these two Acme companies into the older one." The agent runs the merge as a dry run first. It reads the survivor and the duplicate, reports names, domains and how many contacts and deals each carries, and waits for your go-ahead before the real merge.
Processes chain these steps without a prompt each time. CRM hygiene and prospect job changes, Prepare for a demo from the last call and Brief every cold call before it starts all read or write HubSpot. If your pipeline runs elsewhere, Salesforce and Pipedrive work the same way.
What the HubSpot connector can't do
- It needs a service key or a private app token that you create. Tulina holds no shared HubSpot key.
- Notes can be added to a record. There is no separate verb for listing a record's notes, and calls, emails, meetings and tasks are not covered.
- Marketing content, campaigns and conversations are out of its reach.
- A merge cannot be undone. Deleting a record archives it to HubSpot's recycle bin.
- A list's type, manual, dynamic or snapshot, is fixed at creation. Membership changes are refused on a dynamic list, whose criteria are edited instead.
- A search returns at most 10,000 results, so a wider question needs narrower filters.
- A private app token is capped at 190 requests per 10 seconds, so a bulk update is paced rather than instant.
- A contact's page views depend on a HubSpot plan and scope that private app tokens have not been seen to pass. The first page, last page and last session are available.
Why through Tulina
The connection was never the hard part.
Most systems now publish their own way for a model to reach them, and that is good: it is what Tulina connects to. The hard part is deciding who may use it, what it is allowed to return, and what the model knows before it asks.
The keys stay in one place
YouTeamOrganisation
Connect HubSpot once. The credential is encrypted in a vault, never held in clear, and never pasted into a chat or a workflow file. A person uses their own key where they have one, their team's where they do not, the organisation's after that, so nobody is blocked and nobody holds more access than they need. Revoke it once and everything running on it stops.
What leaves is filtered before it leaves
fields droppedmasked or hashedpseudonymised
Roles decide who can reach HubSpot at all. Beyond that, a policy on the connector itself decides what comes back. The data is minimised at the source rather than after it has already reached a model.
Every call is on the record
whatwhowhen
Runs are logged for the organisation. That is what makes an automated action reviewable after the fact, and it is what a security review asks for first.
The run starts knowing your company
namingkey accountsdecisions
A direct connection gives a model access. It does not tell it how your company names things, which accounts matter, or what was decided last quarter. Tulina carries that across every tool, so the same request against HubSpot returns something usable instead of something generic.
Processes that run on HubSpot
Brief every cold call before it starts
A cold calling script opener for every account on the list, written from a dated signal and the lead's public words, saved on the record before anyone dials.
5 steps →Build a target list from a sentence
AI lead generation from one sentence: companies sourced and scored on dated signals, contacts verified, checked against your CRM and staged as a campaign.
6 steps →Catch churn signals and answer them
Usage falling away shows up in analytics before anyone notices on a call, and the drop is read against the account's history into one reactivation email.
4 steps →CRM hygiene and prospect job changes
Weekly CRM data cleansing: duplicates merged with their history kept, prospects who changed jobs moved to their new company, and a recap posted to the team.
4 steps →Draft the monthly investor update
Revenue, pipeline, the plan's own targets, and last month's commitments are pulled together into a draft with the numbers already in it.
5 steps →Expand accounts from usage
Each morning, accounts near a plan limit or on a feature their plan lacks get an offer priced on your packaging rules, an expansion deal and a drafted proposal.
7 steps →Find buyers engaging with competitors
Everyone who engages with a competitor's posts gets found, scored against your ICP, enriched, and staged for outreach with a human approving the list.
5 steps →Follow champions who change jobs
Each week, champions at won accounts are checked for a new employer, the new company is scored on your ICP, and a warm deal opens with a note for the right rep.
7 steps →Follow through on sales calls
After every sales call, the follow-up email is drafted from what was said, the deal gets a dated next step, product asks are filed and any risk is flagged.
7 steps →Outbound from buying signals
Each morning, companies with a new stack or hiring signal are scored on your rubric, checked against the CRM and live sequences, and given one opener per buyer.
7 steps →Prepare for a demo from the last call
The last calls, the deal's state and fresh research become one prep brief, plus a deck built on that company's own use cases, posted before the demo.
5 steps →Social listening to qualified leads
Every morning, competitor-post engagers and companies reading your site are matched to the CRM, qualified on your ICP, ranked by intent and handed to their rep.
7 steps →Turn signups into pipeline
PQL scoring for every trial signup: enriched, scored on your ICP, the first session read, a play chosen by your rules and a deal opened for the right rep.
7 steps →
