SNCF

@sncf what trains leave from Gare de Lyon to Marseille Saint-Charles tomorrow morning between 8 and 10 AM?

Based on the SNCF schedule, here are three direct TGV options departing Gare de Lyon tomorrow morning: TGV 6173 at 8:01 (arrives 11:19, 3h18m), TGV 6107 at 8:41 (arrives 12:01, 3h20m), and TGV 6175 at 9:31 (arrives 12:49, 3h18m). All three are direct services with no changes required. Real-time data shows TGV 6173 is currently operating on schedule. Fares and seat availability would need to be checked separately on the SNCF booking site.

This example demonstrates journey planning between two major stations. The MCP requires valid station IDs (obtained via search) and returns schedule data only—it cannot book tickets or check seat availability. API key must have access to the SNCF coverage region (typically 'fr-idf' for Île-de-France or national coverage).

@sncf are there any service disruptions affecting RER B right now?

Current SNCF service status for RER B shows one active disruption: signal equipment failure at Gare du Nord affecting northbound service between 14:20 and 16:45 today. Trains are running with 10-15 minute delays on the northern branch (toward Aéroport Charles de Gaulle). Southbound service and the Robinson/Saint-Rémy branches are operating normally. The disruption report was last updated 8 minutes ago and includes affected vehicle journeys 847B-23 through 847B-29.

This read-heavy example shows real-time disruption monitoring for a specific line. The MCP returns structured disruption data including severity, time windows, and affected services. Useful for status dashboards, but disruption data is paginated—large incidents may require multiple calls with next_cursor to see all affected journeys.

@sncf find all stations within 2km of coordinates 48.8566, 2.3522 and summarize which lines serve each one

Found 6 SNCF stations within 2km of those coordinates (central Paris): Châtelet-Les Halles serves RER A, B, D and Métro lines 1, 4, 7, 11, 14; Gare de Lyon serves RER A, D, Métro 1, 14 and mainline TGV/Intercités; Saint-Michel Notre-Dame serves RER B, C and Métro 4; Gare du Nord serves RER B, D, E, Métro 4, 5 and mainline services; Gare de l'Est serves Métro 4, 5, 7 and mainline; Réaumur-Sébastopol serves Métro 3, 4. This cluster represents one of Europe's densest transit hubs with connections to regional, national, and international rail.

This synthesis example pairs geographic search with AI reasoning to create a human-readable summary. The MCP's Find Nearby Places tool returns raw station objects; the AI interprets line associations and connectivity. Demonstrates how to combine SNCF data with contextual analysis, though line-service details depend on the coverage region's data completeness.

When your team visits customer sites across French regions

A 6-person field-sales team covering Paris, Lyon, and Marseille needs to book same-day train travel between client meetings. This MCP wins when you're planning routes in Slack or a shared workspace and want realtime departure boards without opening the SNCF app. The journey planner returns station IDs and departure times your team can act on immediately. The service-status tool flags strikes or delays before anyone leaves the office. Trade-off: if your team rarely travels or books trips weeks ahead, the API key setup isn't worth it—just use the consumer site. But for teams doing 10-plus trips a month with tight windows between meetings, this MCP turns route research into a 30-second Switchy query instead of a 5-minute tab-switching exercise.

When you're triaging user complaints about French rail data

A 3-person support team at a European travel aggregator fields 20-30 tickets daily about missing SNCF connections or wrong departure times. This MCP is the right call when your agents need to verify what the SNCF API actually returned versus what your app displayed. The get-transport-object and station-board tools let you pull canonical station data and realtime schedules in the same workspace where you're reading Zendesk tickets. The coverage tool shows the production horizon so you know if a user's complaint is about stale data or a bug in your ingestion pipeline. Threshold: if you're only checking SNCF data once a week, the API key overhead isn't justified. But if you're debugging user-reported discrepancies multiple times per shift, this MCP collapses a 3-step process—find the ticket, open the SNCF docs, run a curl command—into one Switchy prompt.

When you're coordinating offsite arrivals across French cities

A 12-person remote startup is planning a quarterly offsite in Bordeaux with attendees coming from 8 different French cities. This MCP is borderline useful: it shines if one person is wrangling everyone's arrival times and needs to check 15 different station boards in a single planning session. The nearby-places and search-places tools help you find the closest station to your Airbnb, and the journey planner gives you departure options to share in a spreadsheet. Trade-off: if your team is small enough that people book their own tickets, or if you're only doing this twice a year, the MCP is overkill—just send everyone a Google Doc with station names. But if you're the logistics lead running this exercise quarterly and you want to avoid opening 10 browser tabs every time, this MCP turns a 45-minute research task into a 10-minute Switchy session.