Skip to content
Book a build

Is it safe to give Claude write access to HubSpot?

Giving Claude write access to HubSpot is reasonably safe if you limit it. HubSpot's connector supports no deletes, acts only within the user's permissions and logs every create and update in the Audit Log. Confirmations aren't guaranteed: set write tools to Needs Approval, because Always allow can write without asking. The real risk is overwritten values. Use a least-privilege user and test in a sandbox first.

On this page

What HubSpot supports natively

HubSpot builds several guardrails into its Claude connector and MCP server (as of October 2026):

GuardrailWhat it does
No deletesHubSpot says the connector doesn’t support delete operations. One exception showed up in our testing (October 2026): the segment tool offered delete and restore for lists
User permissionsClaude sees and changes only what the connecting user can
Admin approvalNon-admins need a Super Admin to approve the connector and choose its data permissions
Needs ApprovalClaude may show proposed changes and ask you to confirm; set write tools to Needs Approval in Claude’s connector settings so it always asks
Audit LogCreates and updates are attributed to both the user and the Claude connector
Sensitive DataCustom Sensitive Data properties are never available; with Sensitive Data on, engagement data is blocked
Small batchesCreates and updates are capped at 10 records at a time

The gaps those guardrails leave

  • Updates overwrite. No documented delete doesn’t mean no damage. A wrong lifecycle stage on 10 records triggers every workflow that watches it.
  • Validations are skipped. HubSpot notes association label validations aren’t applied to connector writes.
  • Confirmation is a setting, not a guarantee. HubSpot says Claude may ask before writing. With write tools set to Always allow, edits may occur without asking.
  • Properties can’t be edited afterwards. A wrong option value on a new property has to be fixed by hand.

How to set up Claude write access to HubSpot safely

  1. Connect with a dedicated, least-privilege user where you can, not a Super Admin.
  2. Approve only the data permissions you need when the connector is approved. Remember users must reconnect after changes.
  3. Start in a sandbox or developer test account.
  4. Set write tools to Needs Approval in the HubSpot connector’s tool settings in Claude, and read the proposed changes before you type Yes.
  5. Check the Audit Log weekly while the team is learning.
  6. Separate design from execution. Let Claude draft; push structural changes through a reviewed process. See whether to trust Claude with your architecture. For new properties, pipelines and workflows, we recommend Jetstack AI Projects as that reviewed process; see how it compares with Claude and the HubSpot MCP.

Where Jetstack AI fits

For structural changes, Projects replaces open-ended write access with approvals: you describe the build in a brief, approve the plan before anything is generated, approve deployment separately, and get a receipt checked by reading the assets back from HubSpot. Workflows are created switched off, and a Project can change only assets it built. Projects is currently in limited release; see what it can build.

Jetstack’s MCP server applies the same idea with scoped tokens: the read-only preset can’t browse or change HubSpot, deploy scopes are separate, and your Jetstack token is never shown to the assistant. The MCP server docs recommend a token per assistant and keeping your client’s approval prompts on. For deployments made through Portal reuse, you can roll back a Portal reuse deployment: Jetstack shows you what will be removed first, then deletes the assets that deployment created in the destination portal. It won’t delete anything another deployment or Project still uses, and it skips assets that were changed after the deployment (how rollback works). See the developer overview.

Sources

Checked October 30, 2026. HubSpot changes its products often; tell us at team@jetstack.ai if something here is out of date.

AI, Claude & MCP

Do it from your own code or AI agent.

The Jetstack AI REST API, MCP server and CLI share API keys and scopes. Create and manage those keys in the app.

Read the developer docs

AI, Claude & MCP

Related questions.

Is it safe to give Claude write access to HubSpot?

Giving Claude write access to HubSpot is reasonably safe if you limit it. HubSpot's connector supports no deletes, acts only within the user's permissions and logs every create and update in the Audit Log. Confirmations aren't guaranteed: set write tools to Needs Approval, because Always allow can write without asking. The real risk is overwritten values. Use a least-privilege user and test in a sandbox first.

Can I give Claude read-only access to HubSpot?

Limit it through the HubSpot user who connects: Claude can only do what that user's permissions allow. A Super Admin also chooses which optional data permissions the connector gets when approving it.

What if Claude updates the wrong records?

Find the changes in the HubSpot Audit Log, where they're attributed to both the user and the Claude connector. Each record's property history shows the earlier values, but HubSpot's one-click Restore only appears for UI edits, imports and workflows, so plan to re-enter them or export the property history and re-import the old values. Workflows the wrong values triggered have already run.

Should I set write tools to Always allow to save time?

Not on a production portal. HubSpot recommends Needs Approval and warns that with Always allow, edits may happen without asking. Always allow is fine in a test account and a bad idea where a mistake reaches customers.