Key benefits
- Tool access through MCP. Tools configured under an AI Agent or LLM Prompt node are available to MCP-compatible clients.
- Direct flow execution. When a tool is called, Cognigy executes the related flow and returns the result through MCP.
- Standard integration. MCP provides a consistent interface for connecting external applications without custom APIs.
Restrictions
- The following tools are available via MCP:
- Regular Tool
- Knowledge Search
- Send Email
- Execute Workflow
- Handover to agent
- Tool calls must finish within the configured timeout. The default timeout is 30 seconds. If you have an on-premises installation, you can change the default value using the
MCP_SERVER_TOOL_CALL_TIMEOUTvariable. - This endpoint supports OAuth 2.0 or no authentication. User-level authentication isn’t supported.
Prerequisites
- A flow that contains an AI Agent node or an LLM Prompt node.
- The AI Agent node or LLM Prompt node in the flow must include at least one Tool node.
- Each Tool node must have a tool ID and a description.
Generic endpoint settings
Learn about the generic endpoint settings on the following pages:Endpoint-specific settings
Setup tab
AI agent selection
Use this section to select the flow and the AI Agent or LLM Prompt node whose tools are exposed via the Model Context Protocol (MCP). External AI clients can discover and call these tools through this endpoint.MCP server
Integration
This section provides the information required to configure MCP clients: the MCP Server URL and a link to the MCP explorer, where you can view all exposed tools, including their names, descriptions, and parameters.Advanced tab
Authentication
How to set up
The diagram shows the complete MCP workflow. First, make sure your flow is set up correctly with an AI Agent or LLM Prompt node and a Tool node, according to the prerequisites. Then, the MCP client connects to the MCP server endpoint and calls a tool. This triggers the flow execution. The result is returned through a session-based connection.Configure in Cognigy
Configure MCP server endpoint
Configure MCP server endpoint
- In the left-side menu of your project, click Deploy > Endpoints.
- On the Endpoints page, click + New Endpoint.
- In the New Endpoint dialog, select the MCP Server endpoint type and click Create endpoint.
- On the Setup tab, configure the following: Endpoint name — enter a unique name. AI Agent Selection — select the flow and the AI Agent node or LLM Prompt node for which you want to expose the tools via MCP. Save changes.
- In the Integration section, copy the URL in the MCP Server URL field.
- In the MCP server section, activate the Enable MCP Server toggle to start the server and allow clients to connect.
- On the Advanced tab, in the Authentication section, select the authentication method for the MCP server from the Authentication Type list:
- No Auth — no authentication is required to access the MCP server. Use this option for internal or trusted environments only. The option is selected by default.
- OAuth 2.0 — activate token-based authentication using OAuth 2.0. Use this option to verify that only authorized clients can access the MCP server. Configure the following fields:
- JWKS URI — the URL of the JSON Web Key Set (JWKS) endpoint used to verify token signatures. You can find this URL in your identity provider’s OpenID Connect discovery document, typically at
/.well-known/openid-configuration. For example,https://your-idp.example.com/.well-known/jwks.json. - Issuer — the expected token issuer. This value must match the
issclaim in the JWT. You can find this value in your identity provider’s app settings, often labeled Issuer URL or Authority. For example,https://your-idp.example.com. - Audience — the intended recipient of the token. This value must match the
audclaim in the JWT. Use a value from your identity provider, such as the client ID or a custom API identifier. For example,cognigy-mcp-client. - Required Scopes — the OAuth scopes that the token must include to be granted access. Define these scopes in your identity provider when configuring the application or API permissions. For example,
cognigy:mcp. - Allowed Subject IDs — optional. Enter a comma-separated list of subject identifiers (
subclaim) permitted to access the server. You can find the subject ID in the JWT payload or in your identity provider’s user or service account settings. For example,user-123, service-account-456. If the field is empty, all subjects are allowed. - Key Refresh Interval (minutes) — optional. Enter how often the JWKS public keys are refreshed, in minutes. Increase this value to reduce the number of requests to your identity provider. Must be a value between
1and1440, for example,120. If you leave this parameter empty, the server uses the default interval of 60 minutes.
- JWKS URI — the URL of the JSON Web Key Set (JWKS) endpoint used to verify token signatures. You can find this URL in your identity provider’s OpenID Connect discovery document, typically at
Configure in the third-party provider
Configure MCP client
Configure MCP client
The configuration depends on the MCP client you use.
- Direct HTTP integration
- Claude Desktop
- Cursor
The MCP server endpoint supports the MCP Streamable HTTP transport. To integrate directly with the endpoint, follow these steps:
-
Send an
initializerequest.Replace{token}with the token from the MCP Server URL. -
The response includes
mcp-session-id: <session-id>. Store the session ID for use in subsequent requests. -
Request a list of available tools via
tools/list:The request returns the list of available tools with their parameters and descriptions. This allows the client to understand how to call each tool correctly. For example, the response may include tools like this: -
Call tools using the
tools/callmethod, providing the tool name and required arguments. For example, to call theunlock_accounttool with the requiredemailargument, send the following request:After a successful tool call, the MCP server returns the result in a structured MCP response. The response contains the output of the executed tool inside theresultfield. For example, if the tool completes successfully, it may return a text message such asResolvedin thecontentarray.
Use cases
Access knowledge bases
Access knowledge bases
External AI clients can use MCP to query Cognigy knowledge bases through the Knowledge Search Tool.
This approach enables real-time retrieval of relevant content, such as product information, policies, or support articles,
directly from connected applications.
Automate business tasks
Automate business tasks
MCP clients can trigger operational workflows, such as CRM lookups, ticket creation, or email notifications.
When a tool is called, the corresponding flow is executed, allowing existing business logic to run
without modification.
Expose internal logic
Expose internal logic
Existing flows can be published as callable tools. This approach allows organizations to reuse and expose
internal automation, decision logic, and agent configurations as standardized tools
available to external systems.
Connect custom applications
Connect custom applications
Custom AI applications, development environments, or third-party systems can integrate with Cognigy
through the MCP. This approach enables structured tool calls and consistent result handling
without building and maintaining custom APIs.
Troubleshooting
MCP compared to A2A
Use the MCP server endpoint to give an external application access to individual tools from a flow. Use the A2A server endpoint to make a flow available to A2A clients as an A2A agent.