HubSpot CRM Automation via Rube MCP

Automate HubSpot CRM workflows including contact/company management, deal pipeline tracking, ticket search, and custom property creation through Composio's HubSpot toolkit.

Prerequisites

• Rube MCP must be connected (RUBESEARCHTOOLS available)
• Active HubSpot connection via RUBEMANAGECONNECTIONS with toolkit hubspot
• Always call RUBESEARCHTOOLS first to get current tool schemas

Setup

Get Rube MCP: Add https://rube.app/mcp as an MCP server in your client configuration. No API keys needed — just add the endpoint and it works.

1. Verify Rube MCP is available by confirming RUBESEARCHTOOLS responds
2. Call RUBEMANAGECONNECTIONS with toolkit hubspot
3. If connection is not ACTIVE, follow the returned auth link to complete HubSpot OAuth
4. Confirm connection status shows ACTIVE before running any workflows

Core Workflows

1. Create and Manage Contacts

When to use: User wants to create new contacts or update existing ones in HubSpot CRM

Tool sequence:

1. HUBSPOTGETACCOUNT_INFO - Verify connection and permissions (Prerequisite)
2. HUBSPOTSEARCHCONTACTSBYCRITERIA - Search for existing contacts to avoid duplicates (Prerequisite)
3. HUBSPOTREADACRMPROPERTYBYNAME - Check property metadata for constrained values (Optional)
4. HUBSPOTCREATECONTACT - Create a single contact (Required)
5. HUBSPOTCREATECONTACTS - Batch create contacts up to 100 (Alternative)

Key parameters:

• HUBSPOTCREATECONTACT: properties object with email, firstname, lastname, phone, company
• HUBSPOTCREATECONTACTS: inputs array of {properties} objects, max 100 per batch
• HUBSPOTSEARCHCONTACTSBYCRITERIA: filterGroups array with {filters: [{propertyName, operator, value}]}, properties array of fields to return

Pitfalls:

• Max 100 records per batch; chunk larger imports
• 400 'Property values were not valid' if using incorrect property names or enum values
• Always search before creating to avoid duplicates
• Auth errors from GETACCOUNTINFO mean all subsequent calls will fail

2. Manage Companies

When to use: User wants to create, search, or update company records

Tool sequence:

1. HUBSPOTSEARCHCOMPANIES - Search existing companies (Prerequisite)
2. HUBSPOTCREATECOMPANIES - Batch create companies, max 100 (Required)
3. HUBSPOTUPDATECOMPANIES - Batch update existing companies (Alternative)
4. HUBSPOTGETCOMPANY - Get single company details (Optional)
5. HUBSPOTBATCHREADCOMPANIESBY_PROPERTIES - Bulk read companies by property values (Optional)

Key parameters:

• HUBSPOTCREATECOMPANIES: inputs array of {properties} objects, max 100
• HUBSPOTSEARCHCOMPANIES: filterGroups, properties, sorts, limit, after (pagination cursor)

Pitfalls:

• Max 100 per batch; chunk larger sets
• Store returned IDs immediately for downstream operations
• Property values must match exact internal names, not display labels

3. Manage Deals and Pipeline

When to use: User wants to search deals, view pipeline stages, or track deal progress

Tool sequence:

1. HUBSPOTRETRIEVEALLPIPELINESFORSPECIFIEDOBJECT_TYPE - Map pipeline and stage IDs/names (Prerequisite)
2. HUBSPOTSEARCHDEALS - Search deals with filters (Required)
3. HUBSPOTRETRIEVEPIPELINE_STAGES - Get stage details for one pipeline (Optional)
4. HUBSPOTRETRIEVEOWNERS - Get owner/rep details (Optional)
5. HUBSPOTGETDEAL - Get single deal details (Optional)
6. HUBSPOTLISTDEALS - List all deals without filters (Fallback)

Key parameters:

• HUBSPOTSEARCHDEALS: filterGroups with filters on pipeline, dealstage, createdate, closedate, hubspotownerid; properties, sorts, limit, after
• HUBSPOTRETRIEVEALLPIPELINESFORSPECIFIEDOBJECT_TYPE: objectType set to 'deals'

Pitfalls:

• Results nested under response.data.results; properties are often strings (amounts, dates)
• Stage IDs may be readable strings or opaque numeric IDs; use label field for display
• Filters must use internal property names (pipeline, dealstage, createdate), not display names
• Paginate via paging.next.after until absent

4. Search and Filter Tickets

When to use: User wants to find support tickets by status, date, or criteria

Tool sequence:

1. HUBSPOTSEARCHTICKETS - Search with filterGroups (Required)
2. HUBSPOTREADALLPROPERTIESFOROBJECTTYPE - Discover available property names (Fallback)
3. HUBSPOTGETTICKET - Get single ticket details (Optional)
4. HUBSPOTGETTICKETS - Bulk fetch tickets by IDs (Optional)

Key parameters:

• HUBSPOTSEARCHTICKETS: filterGroups, properties (only listed fields are returned), sorts, limit, after

Pitfalls:

• Incorrect propertyName/operator returns zero results without errors
• Date filtering may require epoch-ms bounds; mixing formats causes mismatches
• Only fields in the properties array are returned; missing ones break downstream logic
• Use READALLPROPERTIES to discover exact internal property names

5. Create and Manage Custom Properties

When to use: User wants to add custom fields to CRM objects

Tool sequence:

1. HUBSPOTREADALLPROPERTIESFOROBJECTTYPE - List existing properties (Prerequisite)
2. HUBSPOTREADPROPERTYGROUPSFOROBJECTTYPE - List property groups (Optional)
3. HUBSPOTCREATEPROPERTYFORSPECIFIEDOBJECTTYPE - Create a single property (Required)
4. HUBSPOTCREATEBATCHOFPROPERTIES - Batch create properties (Alternative)
5. HUBSPOTUPDATESPECIFICCRMPROPERTY - Update existing property definition (Optional)

Key parameters:

• HUBSPOTCREATEPROPERTYFORSPECIFIEDOBJECTTYPE: objectType, name, label, type (string/number/date/enumeration), fieldType, groupName, options (for enumerations)

Pitfalls:

• Property names are immutable after creation; choose carefully
• Enumeration options must be pre-defined with value and label
• Group must exist before assigning properties to it

Common Patterns

ID Resolution

• Property display name → internal name: Use HUBSPOTREADALLPROPERTIESFOROBJECTTYPE
• Pipeline name → pipeline ID: Use HUBSPOTRETRIEVEALLPIPELINESFORSPECIFIEDOBJECT_TYPE
• Stage name → stage ID: Extract from pipeline stages response
• Owner name → owner ID: Use HUBSPOTRETRIEVEOWNERS

Pagination

• Search endpoints use cursor-based pagination
• Follow paging.next.after until absent
• Typical limit: 100 records per page
• Pass after value from previous response to get next page

Batch Operations

• Most create/update endpoints support batching with max 100 records per call
• For larger datasets, chunk into groups of 100
• Store returned IDs from each batch before proceeding
• Use batch endpoints (CREATECONTACTS, CREATECOMPANIES, UPDATE_COMPANIES) instead of single-record endpoints for efficiency

Known Pitfalls

• Property names: All search/filter endpoints use internal property names, NOT display labels. Always call READALLPROPERTIESFOROBJECT_TYPE to discover correct names
• Batch limits: Max 100 records per batch operation. Larger sets must be chunked
• Response structure: Search results are nested under response.data.results with properties as string values
• Date formats: Date properties may be epoch-ms or ISO strings depending on endpoint. Parse defensively
• Immutable names: Property names cannot be changed after creation. Plan naming conventions carefully
• Cursor pagination: Use paging.next.after cursor, not page numbers. Continue until after is absent
• Duplicate prevention: Always search before creating contacts/companies to avoid duplicates
• Auth verification: Run HUBSPOTGETACCOUNT_INFO first; auth failures cascade to all subsequent calls

Quick Reference

When to Use

This skill is applicable to execute the workflow or actions described in the overview.

Example

User request:

Automate HubSpot CRM operations (contacts, companies, deals, tickets, properties) via Rube MCP using Composio integration.

Limitations

• Use this skill only when the task clearly matches the scope described above.
• Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
• Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.