Tines MCP Server – AI Access to Workflow Automation
Tines lets teams build custom remote MCP servers directly from a storyboard using the MCP Server action template. The server exposes configured Tines tools to MCP clients so AI agents can invoke approved workflows, APIs, and automations with Tines access controls, credentials, logging, and operational limits in place.
Overview
Tines provides an MCP Server action template for building custom remote Model Context Protocol servers directly on a Tines storyboard. Rather than acting as a connector to one fixed service, a Tines MCP server turns selected Tines workflow capabilities into MCP tools that an AI client can call. This is especially useful for security operations, IT operations, and automation teams that already use Tines to orchestrate investigations, enrichment, notifications, approvals, and remediation workflows.
What the MCP server enables
A Tines MCP server can expose configured tools from a story. Tines documents support for template tools, send-to-story tools, and custom tools built from action groups. Tool names, descriptions, and arguments are made available to the MCP client after a successful connection, so an AI agent can discover the approved actions and invoke them through the server.
The connected tools can call APIs, trigger story logic, enrich data, or route work through Tines actions. The MCP client does not receive direct access to the credentials embedded in those tools; Tines keeps credentials inside the workflow environment and exposes only the configured tool interface.
When to use it
Use the Tines MCP Server when an AI assistant should safely trigger operational workflows rather than receive unrestricted credentials or direct API access. Practical examples include allowing an analyst assistant to enrich an IP address, open a security case, send an alert to a team channel, run a lookup against an internal system, or hand off structured inputs to a Tines story for approval and execution.
It is also useful when teams want to package existing automations as reusable AI-callable tools while retaining Tines governance, logging, access controls, and story design patterns.
Connection and authentication
Tines supports remote MCP servers over Streamable HTTP. The documented server URL format is https://<tenant>.tines.com/mcp/<mcp-path>. Authentication can be configured through the HTTP Authorization header or, for secret-based access, through the URL. Access control modes include public path access, secret-based access, Tines API key access for tenant members or selected teams, and OAuth-based access for tenant members or selected teams.
Tines documents that only Streamable HTTP is supported directly. For clients that only support local stdio servers, mcp-remote can be used as an adapter, but the Tines server itself is still a remote Streamable HTTP MCP server.
Key considerations
Tines MCP servers support tools only; prompts, resources, notifications, sessions, pagination, OAuth as an MCP protocol feature, and JSON-RPC batching are not supported. Tool responses are returned as text. The supported MCP protocol version is 2025-06-18, with partial support for 2025-03-26, while 2024-11-05 is not supported.
Tool response time cannot exceed 30 seconds. Tines also applies concurrency limits shared with response-enabled webhooks, and tool execution is handled asynchronously through the Tines orchestration system. Teams should expose only narrowly scoped tools, configure access control deliberately, and avoid publishing sensitive or destructive workflows without review and approval controls.
Supported Transports
streamable_http
URL: https://<tenant>.tines.com/mcp/<mcp-path>
Frequently Asked Questions
- When should an AI agent use the Tines MCP Server?
- Use it when an AI agent needs to invoke approved Tines workflows, templates, send-to-story tools, or custom action-group tools for security operations, IT operations, enrichment, notification, case handling, or remediation tasks.
- What does the Tines MCP Server add to an AI agent's capabilities?
- It gives the agent a structured MCP tool interface to workflows designed in Tines, allowing the agent to call approved automations without receiving the underlying credentials used by those workflow actions.
- What can the AI agent access or manage through this server?
- The agent can access only the tools configured on that MCP Server action, such as public templates, private templates, send-to-story tools, and custom tools. Tines documents that prompts and resources are not supported.
- How is authentication configured for this MCP server?
- Tines MCP servers can use access controls including path-only access, secret-based access through the URL or HTTP Authorization header, Tines API key access, and OAuth-based tenant or team access. Secrets should be passed as environment variables or client-managed credentials, not hard-coded.
- Which transport should be used for this MCP server?
- Use Streamable HTTP directly with the Tines MCP URL. Tines documents that SSE and native stdio are not supported for the server; clients that only support local stdio can use mcp-remote as an adapter to the Streamable HTTP endpoint.