AI agents are becoming part of everyday business workflows, but they need the right context to work with CRM data.
They need to understand your workspace, available modules and fields, customer information, and the actions they are allowed to perform.
Every Salesmate workspace is different. That makes it difficult for an AI application to work with CRM data if it has to guess how your workspace is configured.
The Salesmate MCP Server solves this by giving compatible AI applications a structured way to discover your Salesmate workspace, access permitted CRM data, and perform supported CRM actions based on the authenticated user's permissions.
What is MCP (Model Context Protocol)?
MCP, or Model Context Protocol, is an open standard that gives AI applications a consistent way to discover and interact with tools and data provided by external systems.
In simple terms, it acts like a common interface between an AI and the software it needs to work with.
For example, an AI assistant connected to Salesmate through MCP can search CRM records, retrieve customer context, update supported records, add notes, or perform other permitted actions.
Technically, MCP defines how an AI application can discover and interact with tools and resources exposed by an MCP server.
So instead of building a separate integration for every AI application, a system can expose its capabilities through MCP in a consistent way.
Think of it this way: APIs let applications talk to each other, while MCP gives AI applications a standardized way to discover and use those application capabilities.
Introduction to Salesmate MCP Server
The Salesmate MCP Server is Salesmate's implementation of the Model Context Protocol (MCP), an open standard that lets compatible AI applications connect with external systems without defining manual inputs and outputs.
It gives AI applications a structured way to understand your Salesmate workspace, access permitted CRM data, and perform supported actions such as searching records, updating information, adding notes, and working with Salesmate Sequences.
How does the Salesmate MCP Server work?
MCP follows a client-server model.
The AI application acts as the MCP client. The Salesmate MCP Server exposes supported Salesmate tools that the application can discover and use.
A typical request looks like this:
User → AI application → Salesmate MCP Server → Salesmate CRM → Result
Suppose a sales manager asks:
"Find the Acme deal and show me its stage, value, owner, and expected close date."
The AI application can determine which Salesmate tools it needs, find the relevant deal, retrieve the requested information, and present the result to the user.
For actions that change CRM data, Salesmate applies the authenticated user's permissions and applicable validation before performing the supported action.
What can you do with the Salesmate MCP Server?
The Salesmate MCP Server provides tools across CRM discovery, records, timelines, notes, and Salesmate Sequences.
The exact capabilities available depend on the authenticated user's permissions, workspace configuration, enabled features, and the current beta scope.
a. Understand your Salesmate workspace
An AI application can discover key workspace information before working with CRM data in Salesmate as a connected sales platform, including:
- The authenticated Salesmate user
- The active workspace
- Available features
- User permissions
- Standard and custom modules
- Module capabilities
- Active users available for assignment or ownership
This helps the application understand what is available in the connected workspace before it searches or changes CRM data, which also helps connect AI workflows to the right CRM context, and the workspace can act as one system for connected CRM context and sales workflows.
b. Discover modules, fields, and valid values
Salesmate workspaces can have different standard and custom modules, fields, layouts, and picklist values. The MCP Server lets an AI application inspect supported CRM metadata, including:
- Available modules
- Module fields
- Field identifiers, labels, and types
- Required and read-only fields
- Searchable fields
- Valid picklist options
- Lookup relationships and record associations
- Create and update permissions
This gives the AI application the context it needs to work with the connected Salesmate workspace instead of guessing field names or values.
c. Search and retrieve CRM records
Connected AI applications can use Salesmate tools to search supported standard and custom modules using:
- Keywords
- Structured filters
- Selected fields
- Sorting
- Owner-based filtering
- Pagination
After finding the right record, the application can retrieve supported contacts, companies, deals, tickets, or custom records.
For example:
"Find Robert Dorf and show me his latest CRM activity."
The AI application can use the relevant Salesmate tools to find the record and retrieve the requested information.
d. Create and update supported records
With the appropriate permissions, an AI agent can use Salesmate tools to create or update supported CRM records, including updating records when permitted.
For example:
"Update the expected close date for the Acme deal."
"Move the Acme deal to the Proposal stage."
"Create a new contact for Robert Dorf."
Before a supported write action is completed, Salesmate can validate information such as:
- Required fields
- Field identifiers and values
- Picklist options
- Owners and associations
- Read-only fields
- Pipeline and stage rules
- Duplicate conditions
- User permissions
The AI application interprets the request and selects the appropriate tool. Salesmate then validates and performs the supported action under the authenticated user's identity, following existing permissions, and successful write actions preserve an audit trail.
e. Review customer timelines
The Salesmate MCP Server can provide structured timeline information across supported activity types, including:
- Emails
- Calls
- Meetings
- Notes
- Activities
- Conversations
- Text messages
- Call transcripts
An AI application can use this information to understand recent customer activity and answer requests such as:
"Summarize the latest calls and emails with Acme."
"What was discussed in our most recent call with Robert?"
"Is there any future activity scheduled for this contact?"
Full email bodies, conversation content, or call transcripts can be requested when needed and when the authenticated user has permission to access them.
f. Add notes to CRM timelines
With the required permissions, an AI application can add a note to a supported Salesmate record.
For example:
"Add a note to the Acme deal that the customer requested revised pricing."
The AI application can identify the relevant record and use the Salesmate note tool to add the information to its timeline.
Work with Salesmate Sequences
The Salesmate MCP Server includes tools for working with supported Salesmate Sequences.
An AI application can:
- Discover available sequences
- Review sequence requirements
- Identify available sender mailboxes
- Identify available sending phone numbers
- Select permitted sender identities
- Enroll eligible contacts into supported sequences
For example:
"Enroll Robert Dorf into the post-demo follow-up sequence."
The AI application can find the contact, locate the relevant sequence, and identify available sender information.
Enrollment remains subject to permissions, sequence availability, sender availability, contact information, unsubscribe status, and other applicable eligibility rules.
Designed for controlled CRM access
The Salesmate MCP Server is designed to give AI applications structured, secure access to Salesmate data and actions while respecting the authenticated user's permissions. This access model also aligns with existing security policies.
a. Permission-aware
The MCP Server respects the permissions of the authenticated Salesmate user. This permission-aware behavior is central to safe data access for AI applications. An AI application cannot use Salesmate MCP tools to view or change information that the user is not permitted to access.
b. Tenant-aware
Each Salesmate workspace can have its own modules, fields, pipelines, users, sequences, and other configurations. The MCP Server can discover supported workspace-specific information instead of relying on a fixed CRM structure.
c. Schema-aware
AI applications can inspect supported module and field information before performing record actions. This helps them work with valid fields, values, required fields, and supported associations.
d. Validated write actions
Supported create and update actions are subject to checks such as user permissions, required fields, valid values, record associations, owners, and pipeline or stage rules.
Together, these capabilities give compatible AI applications the CRM context they need to work with Salesmate in a more controlled and structured way.
Available tools in the Salesmate MCP Server
The initial beta scope includes tools for AI assistants and AI tools to:
- Retrieve current user and workspace context
- Discover standard and custom modules
- Review module schemas
- Find active Salesmate users
- Search and retrieve records
- Create supported records
- Update supported records
- Review record timelines
- Retrieve permitted call transcripts
- Post notes
- Discover available sequences
- Identify available sequence senders
- Enroll contacts into supported sequences
As the beta evolves, this supported scope may also extend to custom tools.
The available capabilities may vary based on workspace configuration, enabled features, permissions, and ongoing beta development.
How is MCP different from the Salesmate API?
MCP does not replace traditional APIs. Both serve different purposes.
Salesmate API: Developers use APIs to build coded integrations and applications that interact with Salesmate programmatically, and API access remains important for developer-built applications.
Salesmate MCP Server: Compatible AI applications can discover supported Salesmate tools and use them based on a user's request, which can reduce the need for custom integrations in some AI application use cases.
In simple terms, APIs give developers a programmatic way to work with Salesmate, while MCP provides a structured tool interface for compatible AI applications, supports natural language interactions and AI-agent workflows, and is distinct from coded integrations.
AI agents and applications: Use cases with Salesmate MCP
The Salesmate MCP Server can support practical CRM workflows across sales and customer management. For example:
Find deals that need attention
A sales manager could ask:
"Which of my open deals need attention?"
The AI application can retrieve relevant deals and review available CRM activity, such as recent interactions, expected close dates, upcoming activities, and next steps, to provide a summary.
Summarize a customer call
A sales representative could ask:
"What was discussed in our latest call with Acme?"
The AI application can use natural language to locate the relevant record, retrieve the latest call and related timeline context, and access the transcript when available and permitted—a real world use case for AI-assisted CRM review.
Update a deal
A user could ask:
"Move the Acme deal to Proposal and update its expected close date."
The AI application can identify the deal and relevant fields, then request the supported update through Salesmate. The action remains subject to the user's permissions and Salesmate's applicable validation.
Enroll a contact in a sequence
A sales representative could use natural language prompts through an MCP client to ask:
"Enroll Robert Dorf into the post-demo follow-up sequence."
The AI application can find the contact and relevant sequence and request enrollment when the applicable requirements and permissions are met, supporting a more efficient outbound workflow without manual intervention.
Build the next generation of AI-powered CRM workflows
AI applications become more useful when they can securely connect to business data and workflows across one sales platform.
The Salesmate MCP Server provides a structured way for compatible AI applications to understand your CRM, retrieve customer context, and perform supported actions. MCP offers a standardized way to enable complex workflows for compatible intelligent agents.
Salesmate provides the CRM tools, workspace context, permissions, validation, and access to supported data.
The connected AI application interprets the user's request, determines which tools to use, analyzes the information returned, and generates the response.
Together, they provide a foundation for AI-powered CRM workflows across sales, customer support, and internal productivity.
The Salesmate MCP Server is currently in beta. Available tools, supported actions, authentication options, and access controls may continue to evolve as the platform develops.
Key takeaways
AI agents are becoming part of everyday business workflows, but they need the right context to work with CRM data.
They need to understand your workspace, available modules and fields, customer information, and the actions they are allowed to perform.
Every Salesmate workspace is different. That makes it difficult for an AI application to work with CRM data if it has to guess how your workspace is configured.
The Salesmate MCP Server solves this by giving compatible AI applications a structured way to discover your Salesmate workspace, access permitted CRM data, and perform supported CRM actions based on the authenticated user's permissions.
What is MCP (Model Context Protocol)?
MCP, or Model Context Protocol, is an open standard that gives AI applications a consistent way to discover and interact with tools and data provided by external systems.
In simple terms, it acts like a common interface between an AI and the software it needs to work with.
For example, an AI assistant connected to Salesmate through MCP can search CRM records, retrieve customer context, update supported records, add notes, or perform other permitted actions.
Technically, MCP defines how an AI application can discover and interact with tools and resources exposed by an MCP server.
So instead of building a separate integration for every AI application, a system can expose its capabilities through MCP in a consistent way.
Think of it this way: APIs let applications talk to each other, while MCP gives AI applications a standardized way to discover and use those application capabilities.
Introduction to Salesmate MCP Server
The Salesmate MCP Server is Salesmate's implementation of the Model Context Protocol (MCP), an open standard that lets compatible AI applications connect with external systems without defining manual inputs and outputs.
It gives AI applications a structured way to understand your Salesmate workspace, access permitted CRM data, and perform supported actions such as searching records, updating information, adding notes, and working with Salesmate Sequences.
How does the Salesmate MCP Server work?
MCP follows a client-server model.
The AI application acts as the MCP client. The Salesmate MCP Server exposes supported Salesmate tools that the application can discover and use.
A typical request looks like this:
User → AI application → Salesmate MCP Server → Salesmate CRM → Result
Suppose a sales manager asks:
The AI application can determine which Salesmate tools it needs, find the relevant deal, retrieve the requested information, and present the result to the user.
For actions that change CRM data, Salesmate applies the authenticated user's permissions and applicable validation before performing the supported action.
What can you do with the Salesmate MCP Server?
The Salesmate MCP Server provides tools across CRM discovery, records, timelines, notes, and Salesmate Sequences.
The exact capabilities available depend on the authenticated user's permissions, workspace configuration, enabled features, and the current beta scope.
a. Understand your Salesmate workspace
An AI application can discover key workspace information before working with CRM data in Salesmate as a connected sales platform, including:
This helps the application understand what is available in the connected workspace before it searches or changes CRM data, which also helps connect AI workflows to the right CRM context, and the workspace can act as one system for connected CRM context and sales workflows.
b. Discover modules, fields, and valid values
Salesmate workspaces can have different standard and custom modules, fields, layouts, and picklist values. The MCP Server lets an AI application inspect supported CRM metadata, including:
This gives the AI application the context it needs to work with the connected Salesmate workspace instead of guessing field names or values.
c. Search and retrieve CRM records
Connected AI applications can use Salesmate tools to search supported standard and custom modules using:
After finding the right record, the application can retrieve supported contacts, companies, deals, tickets, or custom records.
For example:
The AI application can use the relevant Salesmate tools to find the record and retrieve the requested information.
d. Create and update supported records
With the appropriate permissions, an AI agent can use Salesmate tools to create or update supported CRM records, including updating records when permitted.
For example:
Before a supported write action is completed, Salesmate can validate information such as:
The AI application interprets the request and selects the appropriate tool. Salesmate then validates and performs the supported action under the authenticated user's identity, following existing permissions, and successful write actions preserve an audit trail.
e. Review customer timelines
The Salesmate MCP Server can provide structured timeline information across supported activity types, including:
An AI application can use this information to understand recent customer activity and answer requests such as:
Full email bodies, conversation content, or call transcripts can be requested when needed and when the authenticated user has permission to access them.
f. Add notes to CRM timelines
With the required permissions, an AI application can add a note to a supported Salesmate record.
For example:
The AI application can identify the relevant record and use the Salesmate note tool to add the information to its timeline.
Work with Salesmate Sequences
The Salesmate MCP Server includes tools for working with supported Salesmate Sequences.
An AI application can:
For example:
The AI application can find the contact, locate the relevant sequence, and identify available sender information.
Enrollment remains subject to permissions, sequence availability, sender availability, contact information, unsubscribe status, and other applicable eligibility rules.
Designed for controlled CRM access
The Salesmate MCP Server is designed to give AI applications structured, secure access to Salesmate data and actions while respecting the authenticated user's permissions. This access model also aligns with existing security policies.
a. Permission-aware
The MCP Server respects the permissions of the authenticated Salesmate user. This permission-aware behavior is central to safe data access for AI applications. An AI application cannot use Salesmate MCP tools to view or change information that the user is not permitted to access.
b. Tenant-aware
Each Salesmate workspace can have its own modules, fields, pipelines, users, sequences, and other configurations. The MCP Server can discover supported workspace-specific information instead of relying on a fixed CRM structure.
c. Schema-aware
AI applications can inspect supported module and field information before performing record actions. This helps them work with valid fields, values, required fields, and supported associations.
d. Validated write actions
Supported create and update actions are subject to checks such as user permissions, required fields, valid values, record associations, owners, and pipeline or stage rules.
Together, these capabilities give compatible AI applications the CRM context they need to work with Salesmate in a more controlled and structured way.
Available tools in the Salesmate MCP Server
The initial beta scope includes tools for AI assistants and AI tools to:
As the beta evolves, this supported scope may also extend to custom tools.
The available capabilities may vary based on workspace configuration, enabled features, permissions, and ongoing beta development.
How is MCP different from the Salesmate API?
MCP does not replace traditional APIs. Both serve different purposes.
Salesmate API: Developers use APIs to build coded integrations and applications that interact with Salesmate programmatically, and API access remains important for developer-built applications.
Salesmate MCP Server: Compatible AI applications can discover supported Salesmate tools and use them based on a user's request, which can reduce the need for custom integrations in some AI application use cases.
In simple terms, APIs give developers a programmatic way to work with Salesmate, while MCP provides a structured tool interface for compatible AI applications, supports natural language interactions and AI-agent workflows, and is distinct from coded integrations.
AI agents and applications: Use cases with Salesmate MCP
The Salesmate MCP Server can support practical CRM workflows across sales and customer management. For example:
Find deals that need attention
A sales manager could ask:
The AI application can retrieve relevant deals and review available CRM activity, such as recent interactions, expected close dates, upcoming activities, and next steps, to provide a summary.
Summarize a customer call
A sales representative could ask:
The AI application can use natural language to locate the relevant record, retrieve the latest call and related timeline context, and access the transcript when available and permitted—a real world use case for AI-assisted CRM review.
Update a deal
A user could ask:
The AI application can identify the deal and relevant fields, then request the supported update through Salesmate. The action remains subject to the user's permissions and Salesmate's applicable validation.
Enroll a contact in a sequence
A sales representative could use natural language prompts through an MCP client to ask:
The AI application can find the contact and relevant sequence and request enrollment when the applicable requirements and permissions are met, supporting a more efficient outbound workflow without manual intervention.
Build the next generation of AI-powered CRM workflows
AI applications become more useful when they can securely connect to business data and workflows across one sales platform.
The Salesmate MCP Server provides a structured way for compatible AI applications to understand your CRM, retrieve customer context, and perform supported actions. MCP offers a standardized way to enable complex workflows for compatible intelligent agents.
Salesmate provides the CRM tools, workspace context, permissions, validation, and access to supported data.
The connected AI application interprets the user's request, determines which tools to use, analyzes the information returned, and generates the response.
Together, they provide a foundation for AI-powered CRM workflows across sales, customer support, and internal productivity.
The Salesmate MCP Server is currently in beta. Available tools, supported actions, authentication options, and access controls may continue to evolve as the platform develops.
Interested in the Salesmate MCP Server?
Learn how to connect and use the Salesmate MCP Server with our support guide, or contact the Salesmate team to learn more.
Frequently asked questions
1. What is the Salesmate MCP Server?
The Salesmate MCP Server is Salesmate's implementation of the Model Context Protocol. It gives compatible AI applications a structured way to discover supported Salesmate capabilities, access permitted CRM information, and perform supported CRM actions.
2. Does the MCP Server respect Salesmate permissions?
Yes. The MCP Server respects the permissions of the authenticated Salesmate user.
The information and actions available to an AI application depend on the user's permissions, workspace configuration, enabled features, and the current beta scope.
3. Can the MCP Server work with custom modules and fields?
The MCP Server can discover supported standard and custom modules and inspect supported module schemas and fields available in the connected Salesmate workspace.
This allows compatible AI applications to work with workspace-specific CRM structures rather than relying only on predefined field names.
4. Can the MCP Server create or update CRM records?
Yes, where supported and permitted. Create and update actions are subject to applicable validation, including required fields, valid values, associations, pipeline and stage rules, and user permissions.
5. Does MCP replace the Salesmate API?
No. The Salesmate API remains the programmatic interface for developers building integrations and applications. The MCP Server provides a structured tool interface that compatible AI applications can discover and use for supported AI-driven workflows.
6. Which AI applications can connect to the Salesmate MCP Server?
The Salesmate MCP Server is designed for applications that support the Model Context Protocol.
Compatibility, configuration requirements, authentication options, and available capabilities may vary by application.
Shivani Tripathi
Shivani TripathiShivani is a passionate writer who found her calling in storytelling and content creation. At Salesmate, she collaborates with a dynamic team of creators to craft impactful narratives around marketing and sales. She has a keen curiosity for new ideas and trends, always eager to learn and share fresh perspectives. Known for her optimism, Shivani believes in turning challenges into opportunities. Outside of work, she enjoys introspection, observing people, and finding inspiration in everyday moments.