Mapbox Geospatial MCP Server
github.com
Mapbox Geospatial MCP Server
1–10 of 16 posts
Re: Mapbox Geospatial MCP Server
#2Good to see other foundational models get access to location. As long as they support MCP they get it. So far it is only Gemini that has “grounding” with Google Maps.
Re: Mapbox Geospatial MCP Server
#3I wonder if Esri will provide a similar MCP offering.
Re: Mapbox Geospatial MCP Server
#4I've been hoping for Mapbox or ESRI to come out with something robust for this!
This is going to be especially useful in vacation planning. :-)
We're already using ChatGPT and other tools to help plan a vacation overseas, but having good mapping data can help an agent figure out things like "does it make sense to go from A to B by car or train?"
With the right metadata, an agent can churn not only on the actual route planning (for a multi-city itinerary), but also things like "near this AirB&B, there are 5 public parking lots, so it'll be easy to park your car, if you do rent one"
I'm excited for this sort of spatial data to help with things like travel planning, or even "sprinkle these 5 errands into my calendar based on where you know i'll be"
Re: Mapbox Geospatial MCP Server
#5I've been hoping for Mapbox or ESRI to come out with something robust for this! This is going to be especially useful in vacation planning. :-) We're already using ChatGPT and other tools to help plan a vacation overseas, but having good mapping data can help an agent figure out things like "does it make sense to go from A to B by car or train?" With the right metadata, an agent can churn not only on the actual route…
ChatGPT is ages behind supporting MCP. If Mapbox gets special treatment like listed others, case might differ. :)
Re: Mapbox Geospatial MCP Server
#6The demo on the page actually shows an interesting and worrying tendency of LLMs: even given a tool that can fuzzily search for a place by name, the LLM thinks it knows that "cal academy" is the California Academy of Sciences, and it passes that into the search function instead of faithfully transmitting the user's input.
It worked fine in the example, but what if there's a new school in my town called "Cal Academy" for short? Is the LLM just going to assume it knows what I'm talking about?
Seems like you'd need a pretty good system prompt for this to force the LLM to suppress its "world knowledge" — which it's been heavily trained to use as much as possible — and defer to its tools.
Re: Mapbox Geospatial MCP Server
#7I've been hoping for Mapbox or ESRI to come out with something robust for this! This is going to be especially useful in vacation planning. :-) We're already using ChatGPT and other tools to help plan a vacation overseas, but having good mapping data can help an agent figure out things like "does it make sense to go from A to B by car or train?" With the right metadata, an agent can churn not only on the actual route…
can you please elaborate on how this will be helpful for vacation planning? maybe i lack imagination.
Re: Mapbox Geospatial MCP Server
#8I've been hoping for Mapbox or ESRI to come out with something robust for this! This is going to be especially useful in vacation planning. :-) We're already using ChatGPT and other tools to help plan a vacation overseas, but having good mapping data can help an agent figure out things like "does it make sense to go from A to B by car or train?" With the right metadata, an agent can churn not only on the actual route…
can you please elaborate on how this will be helpful for vacation planning? maybe i lack imagination.
Say you want to visit some places for sure and maybe try some restaurant that was recommended to you. You yeet your travel dates to model, it uses mapbox's mcp to find all POIs that you need and creates a detailed plan for you.
Re: Mapbox Geospatial MCP Server
#9The demo on the page actually shows an interesting and worrying tendency of LLMs: even given a tool that can fuzzily search for a place by name, the LLM thinks it knows that "cal academy" is the California Academy of Sciences, and it passes that into the search function instead of faithfully transmitting the user's input. It worked fine in the example, but what if there's a new school in my town called "Cal Academy"…
You don't think the LLM made the leap to California Academy of Sciences specifically because Coit Tower was in the context?
If you asked for driving directions from your local walmart to cal academy, the LLM seems just as likely to decide you mean something else by "cal academy" and use the tools available to determine what.
ETA: Out of curiosity, I tested this on ChatGPT. It did ultimately give directions to California Academy of Sciences. However, during research, it also checked the website of the local school district as well as Waze to determine if I meant other nearby schools, even looking up directions for a local acting school and daycare (each with "Academy" in the name).
Re: Mapbox Geospatial MCP Server
#10Earlier quoted context omitted.
can you please elaborate on how this will be helpful for vacation planning? maybe i lack imagination.
Say you want to visit some places for sure and maybe try some restaurant that was recommended to you. You yeet your travel dates to model, it uses mapbox's mcp to find all POIs that you need and creates a detailed plan for you.
and then you show up to the wrong address because the model swapped the address for a top golf and you wanted hot pot