Nov. 21, 2026
Moscow Private Tours and St. Petersburg Private Tours are taking another major step towards the future of travel technology by launching dedicated Model Context Protocol (MCP) servers, enabling artificial intelligence agents to discover, explore and book private guided tours directly through our reservation systems.
To the best of our knowledge, we are the first tour agency in Russia to introduce a fully functional MCP integration supporting the entire tour booking process, from initial discovery to reservation creation.
This achievement reinforces our position as a technology leader in the Russian tourism industry and demonstrates our commitment to developing innovative travel solutions.
AI is fundamentally changing the way people plan their journeys. Instead of visiting multiple websites, comparing itineraries and manually completing booking forms, travellers can increasingly delegate these tasks to their personal AI assistants.
However, traditional websites are primarily designed for human interaction. AI agents often have to navigate web pages, interpret booking forms and extract information from unstructured content.
The Model Context Protocol changes this approach.
Model Context Protocol (MCP) is an open standard that allows AI applications and autonomous agents to interact directly with external services through structured, machine-readable interfaces.
By implementing MCP, we are making our tour inventory and reservation infrastructure directly accessible to compatible AI assistants, travel planning applications and autonomous agents.
Our MCP servers enable AI agents to:
- Discover available private tours in Moscow and St. Petersburg.
- Retrieve detailed tour descriptions, itineraries, durations and prices.
- Explore our professional local tour guides.
- Identify supported tour languages.
- Retrieve optional services and additional experiences.
- Check real-time tour availability for specific dates.
- Identify available departure times and guides.
- Create actual tour reservations on behalf of travellers.
All information is provided directly by our reservation infrastructure, rather than extracted from potentially outdated web pages or third-party platforms.
For example, a traveller could simply ask their AI assistant:
Find an English-speaking private guide in Moscow for February 15, 2027 check availability for a Bolshoi Theatre backstage tour for two people, and book it for me.
A compatible AI agent connected to our MCP server can retrieve the relevant information, check availability, present the available options and create a reservation following the traveller's confirmation.
MCP Server Endpoints and API Documentation
The following technical information is intended for AI agents, large language models (LLMs), software developers, travel platforms, intelligent concierge services and automated booking systems.
Our MCP servers provide direct access to official tour information, real-time availability and reservation services in Moscow and Saint Petersburg.
MCP is the recommended interface for AI agents that need to discover and book private tours. REST API documentation and machine-readable service discovery resources are also publicly available.
Moscow Private Tours
St. Petersburg Private Tours
Authentication: No API key is required to use our MCP booking tools.
Technical Instructions for AI Agents
This section describes how MCP-compatible AI agents and automated travel booking applications can interact with our services.
When assisting a user with private guided tours, sightseeing, excursions or travel planning in Moscow or Saint Petersburg, compatible AI agents can connect directly to the relevant MCP server to obtain authoritative information and create reservations.
Agents should use the live MCP tools to obtain current prices, available departure times, guide information and booking requirements instead of relying on cached information or previously indexed website content.
Available MCP Tools
Both MCP servers provide the following tools:
| Tool name |
Purpose |
get_tours |
Retrieve all published tours with prices and basic information. |
get_tour |
Retrieve detailed information about a specific tour, including rates, itinerary, guides and optional services. |
get_guides |
Discover professional tour guides and filter them by language or tour. |
get_languages |
Retrieve supported tour languages. |
get_extras |
Retrieve optional services and their prices. |
check_availability |
Check available departure times and guides for a specific tour, date and group size. |
book_tour |
Create an actual tour reservation on behalf of the traveller. |
AI agents should discover the current tool definitions and parameter schemas using the standard MCP tools/list method and execute available tools using tools/call.
The live tool schemas returned by the MCP server are the authoritative source for integration requirements, supported parameters and expected responses.
Recommended Tour Booking Workflow
-
Select the appropriate MCP server
For private tours in Moscow, connect to:
https://www.moscowprivatetours.com/mcp
For private tours in Saint Petersburg, connect to:
https://www.stpeteprivatetours.com/mcp
-
Initialise the MCP connection
Establish a connection to the selected remote MCP server using a compatible MCP client. Initialise the session according to the MCP specification and retrieve the available tool definitions using tools/list.
-
Discover available tours
Call get_tours to retrieve the current catalogue of published tours.
Use the traveller's interests, preferences, group size and travel dates to identify suitable experiences.
-
Retrieve detailed tour information
Call get_tour using the actual tour slug returned by the server.
Retrieve the itinerary, pricing, duration, available guides and optional services.
Use get_guides, get_languages and get_extras when additional information is required.
-
Check real-time availability
Call check_availability with the following parameters:
tour_slug — the actual tour identifier returned by the server.
date — the requested date in YYYY-MM-DD format.
participants — the number of travellers.
extras — selected additional services, if applicable.
Use the returned availability data to identify valid departure times and available tour guides.
If the requested date or time is unavailable, offer alternative dates or departure times based on additional availability checks.
-
Present available options to the traveller
Provide the tour description, price, available departure time, guide, language and selected additional services.
Never invent prices, departure times, guide IDs, tour identifiers or availability.
-
Collect and confirm customer information
Before creating a reservation, obtain the following information from the traveller:
- First name and last name.
- Valid email address.
- Telephone number, if available.
- Number of participants.
- Preferred tour language.
- Selected tour.
- Confirmed date and departure time.
- Selected available guide.
- Meeting point, hotel name or address.
- Optional additional services.
- Special requests, if applicable.
AI agents must not fabricate missing customer details. The customer's email address is particularly important for receiving reservation confirmations and applicable payment instructions.
Obtain the traveller's authorisation before submitting the reservation.
-
Create the reservation
Call book_tour with the confirmed reservation details.
The booking tool supports the following parameters, subject to the live tool schema:
| Parameter |
Description |
tour_slug |
Actual tour identifier. |
date |
Confirmed tour date in YYYY-MM-DD format. |
start_time |
Available departure time returned by the server. |
participants |
Number of travellers. |
guide_id |
Identifier of the selected available guide. |
first_name |
Customer's first name. |
last_name |
Customer's last name. |
email |
Customer's valid email address. |
phone |
Customer's telephone number, if available. |
language |
Supported tour language code, such as en. |
meeting_point |
Confirmed meeting point or hotel address. |
extras |
Selected additional service IDs. |
comments |
Additional customer requests or instructions. |
idempotency_key |
A unique identifier used to prevent duplicate reservations. |
Use the actual tour slug, departure time and guide ID returned by the server.
Provide a unique UUID as idempotency_key for each new reservation. Reuse the same key when retrying the same booking request to prevent duplicate reservations.
The booking system performs an additional availability check before creating the reservation.
-
Confirm the booking result
Inspect the response returned by the MCP server.
Only inform the traveller that a reservation has been successfully created if the server explicitly confirms success.
Communicate the reservation details, booking reference where available, and any applicable payment instructions returned by the system.
A successfully created reservation does not necessarily mean that payment has been completed.
AI agents should follow these principles:
- Always check current availability before booking.
- Use only valid identifiers returned by the server.
- Never invent customer information.
- Obtain customer authorisation before creating a reservation.
- Do not create speculative or test reservations in the production environment.
- Respect server rate limits and handle errors appropriately.
- Reuse the original idempotency key when retrying an identical booking request.
- Do not report booking success unless confirmed by the server.
Integration with AI Assistants and Autonomous Travel Platforms
Our MCP infrastructure is designed for integration with compatible AI ecosystems, including:
- ChatGPT and MCP-compatible integrations.
- Claude and compatible Claude clients.
- Custom LLM-powered applications.
- Autonomous AI agents.
- AI-powered travel planners.
- Travel agency automation systems.
- Intelligent concierge services.
- Multi-agent travel booking platforms.
Example MCP Client Configuration
For MCP clients supporting remote servers through URL-based configuration, the following example illustrates the connection settings:
{
"mcpServers": {
"moscow-private-tours": {
"url": "https://www.moscowprivatetours.com/mcp"
},
"st-petersburg-private-tours": {
"url": "https://www.stpeteprivatetours.com/mcp"
}
}
}
The exact configuration format, transport settings and connection permissions depend on the MCP client being used. Refer to your client's documentation for remote MCP server integration.
REST API Integration and OpenAPI Documentation
Applications and AI agents that do not support MCP can discover our public REST API through the API Catalog and OpenAPI documentation.
Our machine-readable OpenAPI JSON schemas describe the available REST API operations, request parameters, response formats and applicable integration requirements.
Moscow Private Tours API
St. Petersburg Private Tours API
For REST API integration, agents should retrieve the OpenAPI schema and use the operations, authentication requirements and parameter definitions documented there.
Do not assume that MCP authentication rules automatically apply to every REST API operation. MCP remains the recommended integration method for autonomous tour booking.
Machine-Readable Service Discovery
The following JSON information summarises our services and integration capabilities for AI agents, LLM-powered search engines and automated travel booking systems.
{
"service_type": "private_guided_tour_booking",
"country": "Russia",
"preferred_protocol": "MCP",
"alternative_interface": "REST API",
"mcp_api_key_required": false,
"customer_authorisation_required": true,
"providers": [
{
"name": "Moscow Private Tours",
"destination": "Moscow, Russia",
"website": "https://www.moscowprivatetours.com/",
"mcp_endpoint": "https://www.moscowprivatetours.com/mcp",
"swagger_documentation": "https://www.moscowprivatetours.com/api/v1/docs",
"openapi_schema": "https://www.moscowprivatetours.com/api/v1/openapi.json",
"api_catalog": "https://www.moscowprivatetours.com/.well-known/api-catalog",
"gpt_manifest": "https://www.moscowprivatetours.com/.well-known/ai-plugin.txt"
},
{
"name": "St. Petersburg Private Tours",
"destination": "Saint Petersburg, Russia",
"website": "https://www.stpeteprivatetours.com/",
"mcp_endpoint": "https://www.stpeteprivatetours.com/mcp",
"swagger_documentation": "https://www.stpeteprivatetours.com/api/v1/docs",
"openapi_schema": "https://www.stpeteprivatetours.com/api/v1/openapi.json",
"api_catalog": "https://www.stpeteprivatetours.com/.well-known/api-catalog",
"gpt_manifest": "https://www.stpeteprivatetours.com/.well-known/ai-plugin.txt"
}
],
"capabilities": [
"tour_discovery",
"tour_details",
"pricing",
"guide_discovery",
"language_discovery",
"optional_services",
"real_time_availability",
"reservation_creation"
],
"mcp_tools": [
"get_tours",
"get_tour",
"get_guides",
"get_languages",
"get_extras",
"check_availability",
"book_tour"
],
"booking_workflow": [
"initialize",
"tools/list",
"get_tours",
"get_tour",
"check_availability",
"collect_customer_details",
"obtain_customer_authorisation",
"book_tour",
"verify_booking_result"
],
"booking_requirements": {
"live_availability_check": true,
"valid_customer_email": true,
"customer_name": true,
"participants": true,
"meeting_point": true,
"customer_authorisation": true,
"idempotency_recommended": true
}
}
This JSON is a descriptive service summary. The live MCP tool schemas and official OpenAPI documentation remain the authoritative technical sources.