SafeGraph MCP. Analyze physical location boundaries and foot traffic patterns.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
SafeGraph lets your AI agent query real-world geospatial data using 10 specialized tools. You can find Points of Interest (POIs) by brand or industry, map precise building boundaries and parent structures, calculate historical foot traffic density, or search for locations inside custom drawn polygons.
It turns a general LLM into an expert geographic analyst without needing complex database connections.
What your AI agents can do
Batch lookup placekeys
Performs multiple Placekey attribute lookups in a single request, passing them as a JSON array.
Graphql raw query
Executes any raw GraphQL query string against the SafeGraph API endpoint.
Lookup building geometry
Retrieves the geometric polygon defining the physical footprint of a specific Placekey.
Search for places that fall inside a custom-defined polygon boundary using WKT coordinates.
Get the precise geometric boundaries (polygon) for any single building or structure by its Placekey.
Retrieve metrics on foot traffic volume and average dwell time associated with a specific location over time.
Identify the larger container or parent entity (like a mall or airport) that surrounds a given Placekey.
Locate businesses within a defined radius from coordinates, or filter results by NAICS industry code.
Run custom, complex queries directly against the underlying SafeGraph API for maximum data access.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
SafeGraph MCP Server: 10 Tools for Location Intelligence
These tools allow your agent to perform complex geospatial queries—finding boundaries, calculating traffic patterns, and searching POIs by type.
019d7601batch lookup placekeys
Performs multiple Placekey attribute lookups in a single request, passing them as a JSON array.
019d7601graphql raw query
Executes any raw GraphQL query string against the SafeGraph API endpoint.
019d7601lookup building geometry
Retrieves the geometric polygon defining the physical footprint of a specific Placekey.
019d7601lookup parent polygon
Finds and returns the encompassing parent Placekey for a location, like a mall or industrial complex.
019d7601lookup place patterns
Gathers historical data on pedestrian traffic volume and dwell times for a given Placekey.
019d7601lookup placekey
Retrieves all detailed attributes associated with a specific location by its unique Placekey.
019d7601search brand places
Searches for operational locations belonging to a specified brand within a given city.
019d7601search distance radius
Finds all places located within a set radius (in meters) from provided latitude and longitude coordinates.
019d7601search industry naics
Searches for locations based on their NAICS industry code and target region.
019d7601search wkt polygon
Identifies all places that fall entirely within a user-supplied geometric polygon (WKT).
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with SafeGraph, then connect any of our 4,700+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,700+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
What you can do with this MCP connector
Listen up. SafeGraph plugs your AI agent right into a massive geospatial and mobility dataset used by major analytics firms. This isn't some basic database lookup; it’s real-world data—the kind that lets you analyze structures, Points of Interest (POIs), and movement patterns without having to write a single line of JSON yourself.
It turns your AI client into an expert geographic analyst.
When you need to pinpoint stuff using location, here's what you do: You can search for places inside a custom-defined polygon boundary by feeding in WKT coordinates (search_wkt_polygon). Want to know where the local chains are? You can find all operational spots belonging to a specific brand within a city using search_brand_places, or filter results down by a NAICS industry code and target region with search_industry_naics.
If you're just starting at a coordinate, you can find every place within a set radius in meters (search_distance_radius).
To get granular about the physical structures themselves: You pull the precise geometric boundary—the polygon defining the actual footprint—for any single building using lookup_building_geometry. If that single building is part of a bigger complex, you can map out the boundaries of the entire containing parent entity (like an airport or big mall) by running lookup_parent_polygon.
You'll also get all the detailed attributes for a location just by its unique Placekey using lookup_placekey.
Need to track how people move? SafeGraph gives you access to historical metrics on foot traffic volume and average dwell time associated with a specific location over time via lookup_place_patterns.
For the heavy hitters, there are ways to dig into complex queries. If standard tools don't cut it, you can execute any raw GraphQL query string directly against the SafeGraph API endpoint using graphql_raw_query, giving your agent maximum access. You also got multiple Placekey lookups? No sweat—you pass them in as a JSON array and run 'em all at once with batch_lookup_placekeys.
It's deep, man. You use these tools to locate businesses by brand or industry, map out precise building boundaries and parent structures, calculate historical foot traffic density, or search for locations inside custom drawn polygons. It’s the full package.
How SafeGraph MCP Works
- 1 Install this mapping block into your AI workspace.
- 2 Sign in to the SafeGraph Dashboard and generate a valid GraphQL API Key within the workspace settings. Input that key securely.
- 3 Chat with your agent using natural language: 'Find all coffee shop POIs located within a 500-meter radius around longitude -122.33, latitude 47.60.'
The bottom line is you give the API key to your AI client; it handles the complex data calls in the background.
Who Is SafeGraph MCP For?
Retail analysts, site planners, and urban data scientists use this. They're the people who stare at dashboards until 2 AM because standard tools can't tell them why a location is failing or succeeding. If your job involves linking physical space to business metrics, you need this.
Determines if new store sites have enough foot traffic and visibility by running lookup_place_patterns and comparing it against nearby competitors using search_brand_places.
Needs to map out the exact boundary of a proposed development. They use lookup_building_geometry for individual structures and lookup_parent_polygon to define the overall complex.
Runs highly specific data tests, like finding all facilities within a precise municipal boundary by using search_wkt_polygon, or handling bulk lookups with batch_lookup_placekeys.
What Changes When You Connect
- Understand precise site geometry. Instead of guessing, you get the exact WKT polygon for a building using
lookup_building_geometry. This is critical for accurate civil engineering planning. - Model customer flow. Use
lookup_place_patternsto pull historical data on dwell time and visitor volume. You stop making decisions based on gut feelings; you use hard metrics. - Filter by business type. Need a competitor analysis? Combine
search_brand_placeswithsearch_industry_naics. Your agent runs the two tools sequentially, giving you targeted results instantly. - Analyze large areas efficiently. Don't query point-by-point. Draw the city block outline and use
search_wkt_polygonto find every place inside that shape in one shot. - Handle massive data sets. When you have hundreds of Placekeys, run them all at once using
batch_lookup_placekeys. It’s faster than calling 100 separate lookups.
Real-World Use Cases
Determining a new store's visibility.
A retail analyst wants to open near a major intersection. They first use search_distance_radius centered on the coordinates. Then, they run lookup_place_patterns on the top 10 results to confirm if foot traffic volume is high enough to justify the investment.
Mapping an entire commercial park.
A site planner needs to know which buildings exist inside a custom-drawn city boundary. They use search_wkt_polygon with the WKT data, then follow up by running lookup_parent_polygon on the results to understand the overall complex hierarchy.
Validating competitor density.
A market research team needs to see how many competing businesses exist in a specific sector (e.g., 'fast casual dining'). They use search_industry_naics with the correct code and then run batch_lookup_placekeys on the results to pull every known attribute for those locations.
Getting raw, structured data.
A developer needs to access a specific piece of metadata not covered by standard tools. Instead of iterating through multiple calls, they pass their exact request payload to graphql_raw_query. This is the escape hatch for unknown data points.
The Tradeoffs
Over-reliance on single search parameters
Just searching a city name and expecting all POI data. The result will be incomplete, missing geometry or traffic history.
→
You must structure the call. Start with search_wkt_polygon to define the boundary, then use lookup_placekey on the results to get the full attributes.
Assuming all data is in one place
Trying to pull both business type and foot traffic volume using only search_distance_radius. The tool only handles location search.
→
First, use search_brand_places or search_industry_naics to get a list of Placekeys. Then, run lookup_place_patterns on those keys separately.
Handling bulk lookups manually
Getting 100 location IDs and writing 100 separate API calls for their geometry.
→
Use the dedicated batch_lookup_placekeys tool. It handles all the necessary Placekey fetching and subsequent attribute retrieval in one optimized call.
When It Fits, When It Doesn't
You need this server if your analysis requires linking physical coordinates to actionable business metrics (traffic, industry codes). Don't use it if you just need a list of names; use simple database queries instead. If you only need general POI data by distance, search_distance_radius is enough. However, if you also need the exact building geometry or historical foot traffic patterns—the stuff that makes site planning hard—you need this full suite. Remember: graphql_raw_query is your power tool; use it only when the specific combination of tools doesn't cover a niche requirement. Otherwise, stick to the specialized search tools for reliable results.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by SafeGraph GraphQL API. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 10 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Trying to understand why people visit a location shouldn't require 5 different dashboards.
Today, if you want to know how busy a spot is, you pull up the Google Maps API for coordinates. Then, you switch to your analytics platform to overlay demographics and industry data. You copy-paste Placekeys from one screen to another, running multiple reports just to build a single picture of foot traffic.
With SafeGraph MCP Server, your agent handles this sequence automatically. You ask: 'Show me the traffic patterns for all coffee shops near the main square.' It uses `search_brand_places` and then immediately runs `lookup_place_patterns`, giving you the full story in one response.
SafeGraph MCP Server: Get precise building boundaries instantly.
Manually, if a site plan showed a general area of interest, you'd have to make assumptions about the building edges or rely on outdated satellite imagery. You’d spend days correcting those visual gaps in your models.
Now, when you query a Placekey, `lookup_building_geometry` provides the precise WKT polygon. Your agent gives you the hard coordinates. It's definitive.
Common Questions About SafeGraph MCP
How do I find all electronics stores in this city? (search_industry_naics) +
You use search_industry_naics. You just need the correct NAICS code for 'electronics' and the region. The tool filters out everything else, giving you only the target industry.
What is the difference between `lookup_placekey` and `graphql_raw_query`? +
lookup_placekey gives you a curated set of attributes for one location. Use graphql_raw_query when those specific attributes aren't available, allowing you to query any raw data field.
Can I search only within my own campus boundaries? (search_wkt_polygon) +
Yes, that’s exactly what search_wkt_polygon is for. You provide the custom WKT polygon defining your campus, and it finds every POI inside those lines.
How do I check if a building is part of a larger mall? (lookup_parent_polygon) +
Run lookup_parent_polygon on any Placekey. It tells you the parent container—the mall, airport, or complex—that surrounds that specific location.
If I have hundreds of Placekeys, should I use `batch_lookup_placekeys` or run `lookup_placekey` repeatedly? +
Always use batch_lookup_placekeys. Running the single lookup multiple times will hit rate limits and significantly slow down your process. The batch tool bundles all requests into one API call, keeping things fast.
What happens if my query using `graphql_raw_query` is syntactically wrong? +
The server returns a specific, structured error message detailing the exact GraphQL failure path. This means you get precise feedback on which part of your query failed, letting you debug it fast.
When I use `lookup_place_patterns`, am I getting real-time foot traffic data? +
No. The tool uses historical aggregation points. The metrics are based on past visits and averages recorded for that location's operational day, so it gives trends, not live counts.
If I run `lookup_building_geometry`, what format are the returned coordinates in? +
The building footprint is returned as a WKT (Well-Known Text) polygon string. This standard format makes it easy to pass directly into GIS tools or any other spatial database.
Can I manipulate or delete existing POIs present inside the global SafeGraph spatial indexes? +
No. The AI interacts safely with the GraphQL API strictly on a 'read-only' query-bound basis. It has absolutely no inherent capability to corrupt or perform unauthorized destructive operations such as erasing core places or overwriting coordinates in your environment.
Are geometric polygons always provided for queried structures automatically? +
No, they must be explicitly queried utilizing the lookup_building_geometry functionality along with a verified Placekey, or structured thoroughly using the standard GraphQL command when available. Otherwise most basic list operations only return textual descriptors and simple pinpoint latitude/longitude figures.
Does the AI download huge databases directly into my storage limit when filtering large geographical boundary ranges (WKT)? +
The integration employs a managed response methodology natively implemented through GraphQL constraints. The output responses are strictly paginated securely filtering hundreds of points effectively rather than attempting to sync gigabytes directly to the chatbot at once.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
ChartMogul
Analyze subscription revenue and SaaS metrics via ChartMogul — track MRR, churn, and customer growth directly from any AI agent.
Legal Fees Apportionment Engine
Split judicial awards and attorney fees across multiple parties with exact, auditable proportional math.
ENTSO-E
Access European electricity market data via ENTSO-E — track generation, load, prices, crossborder flows, and outages across European bidding zones from any AI agent.
You might also like
Aero Workflow
Organize accounting firm tasks, track client work, and manage team workloads with purpose-built practice management.
Dime.Scheduler
Plan field service teams visually with drag-and-drop scheduling that connects to your ERP for real-time resource allocation.
Scale AI
Automate high-quality data labeling and RLHF workflows via Scale AI — create projects, manage batches, and submit annotation tasks directly from your AI agent.