FutureSearch MCP Server

The tool requires FUTURESEARCH_API_KEY, which is a sensitive credential; if exposed via client logs, exported environment variables, or misconfiguration, it could be abused to access the associated service. The provided materials do not show requests for additional secrets or any obvious credential-abuse design.

Although no remote host is listed, the tool requires an API key and its 'research/forecast/classify dataset' description reasonably suggests communication with FutureSearch services and possible transmission of user-submitted data. Because the README is missing and endpoints are undisclosed, destination and scope lack transparency, but there is no explicit red flag indicating exfiltration to unrelated or suspicious endpoints.

System checks indicate that this MCP executes code / runs a local process, which is normal for this type of tool. The available materials do not show requests for system privileges beyond its stated function, nor explicit signs of arbitrary script execution, privilege escalation, or persistence.

The description indicates it processes every row of a dataset, so it will likely access user-provided structured data. The materials do not specify what local files, databases, or other resources it can read or write; with no README, the access boundary is unclear, but there is also no evidence of permissions clearly excessive for its stated function.

It comes from an official registry, has an auditable open-source repository, and was updated within the last year, all of which are strong positive signals. The missing license and low GitHub star count suggest limited legal/ecosystem maturity, but based on the available facts there is no supply-chain red flag such as closed-source distribution, abandonment, or suspicious provenance.