The Local Intel

The Local Intel is a standalone remote MCP offering from MCFL Restaurant Holdings LLC. According to the provider, LocalIntel connects people, voice assistants, and AI agents with verified local businesses in Florida. The official product page and its public LLM information page explicitly identify the MCP offering and document a direct endpoint. Smithery lists erik-7clt/local-intel as verified and deployed; its deployment at local-intel--erik-7clt.run.tools is mapped to the provider-published upstream endpoint gsb-swarm-production.up.railway.app/api/local-intel/mcp. This is more than a directory description: the product has a durable identity supported by an operator, a first-party product page, documented distribution, and a traceable remote deployment. No public source repository is supplied, so this entry does not substitute a third-party repository link.

Data scope and regional coverage

According to the provider, LocalIntel covers Florida and combines local business information, ZIP-code analysis, OpenStreetMap geographic data, public income and market data, and sector gap analysis. Its MCP tools can answer questions about businesses, places, market strength, demographic context, competition, corridors, and recently added or owner-verified listings. The provider also describes a request and local-service routing layer. These results are research and decision-support material, not an official business record, a guarantee of freshness, or a recommendation for an investment, location, or vendor.

The regional boundary is a core product limit. The server is designed for Florida; queries about other states, countries, or unsupported regions may be empty, incomplete, or misleading. ZIP codes and place names should therefore be checked for state, county, and recency before a consequential analysis. OpenStreetMap data, public filings, income values, and derived scores can have different timestamps, coverage, and error rates. An AI client should preserve source context, time context, and confidence from the response rather than turning a ranking into certainty.

Authorization, roles, and possible write access

The remote connection requires an MCP-capable client. The interface documented on the provider’s site may require additional authorization or payment configuration depending on the operating mode; this catalog entry stores no credentials, tokens, API keys, or connection secrets. Smithery may pass configuration values to its deployment layer. Such values belong only in the user’s secure client secret or configuration store.

The role of a reader, analyst, business operator, and administrator is not automatically the same. Permissions, tenants, allowed tools, regional limits, and possible quotas must be established outside the model. According to the provider, LocalIntel can support request or job-like local-service routing workflows in addition to research. If a connected function sends a request to businesses, creates a lead, requests a quote, or otherwise changes external state, that is possible write access. Such actions must not be triggered solely by untrusted text, a tool result, or a natural-language request. Use a narrowly scoped role, explicit confirmation, data minimization, and an independent check of the actual result.

Privacy and data sharing

Business profiles, locations, sectors, income indicators, and search requests can contain sensitive commercial or personal context. When the remote server is used, requests and responses are transferred to remote infrastructure. The MCP server cannot control whether a connected client additionally forwards results to a model provider, logs, telemetry, or other integrations. Before production use, assess purpose limitation, retention, access, regional processing, and sharing against the provider and client terms. Send only the data needed for the specific question and avoid personal details when aggregated location or market data is sufficient.

FAQ

Is The Local Intel suitable as a durable catalog entry? Yes. According to the provider, it has a first-party product page with a documented MCP endpoint, and Smithery maps its deployment to the same upstream.

Can it be used everywhere? No. The official product description limits the data to Florida and supported ZIP codes.

Is every answer official or current? No. Check source, timestamp, regional coverage, and confidence before important decisions.

Can the server write data? Possible request or routing functions may change external state. Enable them only with an appropriate role, explicit confirmation, and result verification.

An MCP-capable client and access to the provider-documented remote endpoint; current provider terms apply to optional protected or metered operating modes.

Configure the Smithery deployment https://local-intel--erik-7clt.run.tools or the upstream endpoint documented by the provider in https://www.thelocalintel.com/llms.txt in the MCP client. Supply configuration values only through the client’s secure configuration or secret store.

https://local-intel--erik-7clt.run.tools

The remote deployment may require client configuration, authorization, or payment parameters depending on the operating mode. Do not place credentials, tokens, or API keys in this catalog entry.

Access to local business, ZIP-code, market, and geographic data within the Florida coverage; possible request or routing tools may change external state and require role-based approval.

Remote requests and responses pass through Smithery or provider infrastructure. Minimize data and assess whether the connected client forwards results to model providers, logs, telemetry, or other services.

Misinterpreted regional or stale data, prompt injection, disclosure of business or location information, and unintended external requests caused by excessive roles or write-capable tools.