Timetta MCP Server

Connecting external AI assistants to Timetta

Connecting external AI assistants

Timetta provides its applied tools to external AI clients via Model Context Protocol (MCP). MCP allows compatible external AI clients to find data and perform allowed operations in Timetta.

Key connection features

  1. Connection address
    For cloud Timetta: https://api.timetta.com/mcp. For on-premises installation: https://<WebAPI-address>/mcp.
  2. Connection method
    It is recommended to use an MCP client. The server uses MCP Streamable HTTP (stateless).
  3. Authorization
    Uses a Bearer token (Timetta API token issued on behalf of the user).
MCP tools

What you can do via MCP

Almost all Timetta functions are available via MCP, allowing an external AI assistant to work with platform data and features.

Available operations

  1. All Timetta features
    Almost all platform functions are available via MCP, enabling an external AI assistant to interact with data and utilize Timetta features.
  2. Read tools
    Include functions for searching employees, tasks, and reading Wiki.
  3. Atomic change tools
    Allow performing individual actions, such as creating tasks or editing Wiki pages.
  4. Complex tools
    Provide capabilities for more complex operations, such as project planning.
MCP capabilities

Security and permissions

MCP operations are performed on behalf of the user who owns the token and respect their access rights. The server checks scopes: mcp: read (read), mcp: write (read and write), all: read, all: write. Scopes do not extend user rights.

Access control

  1. Scopes
    Define permissions for operations: read, read and write, etc.
  2. Permission checking
    The server takes into account the user rights associated with the token.
MCP security

Compatibility with the built-in assistant

The public MCP interface and the built-in AI assistant's internal tool catalog evolve independently, and their contents may differ. Data-changing tools should be called only at the user's request.

Compatibility features

  1. Independent development
    The MCP interface and the built-in assistant develop separately.
  2. Tool invocation
    Data modification is possible only at the user's request.
MCP compatibility

Frequently asked questions

Answers to common questions about Timetta MCP Server

MCP is designed for connecting external AI clients, while the built-in assistant operates within the Timetta interface.

Operations are performed on behalf of the user, taking into account their access rights. Bearer token is used for authorization, and the server checks scopes.

Supports mcp: read (read), mcp: write (read and write), all: read, all: write. Scopes do not extend user rights.

Almost all Timetta functions are available via MCP: read tools (search for employees, tasks, read Wiki), atomic change tools (create and edit tasks, Wiki pages), and complex tools (e.g., project planning).

Yes, for on-premises installation, use the address https://<WebAPI-address>/mcp.

You can get a list of tools via the tools/list operation. Availability depends on the token scopes and user rights.

Connect Timetta to your AI assistant

Start your free 14-day trial and try out Timetta’s features for yourself.

AI

Похоже, вам удобнее русский язык. Перейти на русскую версию?