Information the connector receives
Your MCP client sends the tool name and arguments needed for a request. These can include search text, company identifiers, industry and task selections, benchmark names, metrics, method filters and pagination settings. The connector queries public company, product, ranking and research records; it does not require an account, API key, private files or payment information.
The connector does not request your full conversation history. Your client decides which text to include in tool arguments, so information from a conversation may reach us if the client includes it in a query. Please use public research and company queries and avoid submitting confidential documents, credentials or sensitive personal information.
Our hosting and network infrastructure processes request metadata, which can include IP addresses, user-agent information, timestamps, request paths, response status and diagnostic information. Operational logging and tracing are enabled. If you contact support, we receive your email address and the information you send.
How information is used
We use requests to retrieve the requested public information, return results to your client, maintain service availability, diagnose failures, prevent abuse and address support or correction requests. The MCP application does not maintain a conversation history or persistent user session. Clients must send the relevant identifiers and filters with each request.
Hosting, service providers and disclosure
The connector runs on Cloudflare infrastructure and uses SOTA2's catalog and research services. Relevant query parameters are passed to these services to retrieve results. Cloudflare may process network and diagnostic information as described in its privacy policy. Infrastructure and service processing may take place in countries other than your own.
Information may be disclosed when necessary to operate and secure the connector, investigate abuse, respond to your support request, or meet applicable legal requirements. Your AI client, including Claude or ChatGPT, separately handles your conversations, connection settings and tool results under its own privacy policy. Following a source or paper link opens a separate website with its own practices.
Storage and retention
The connector does not create a user profile or store conversations in an application database. Public query results may be cached. Infrastructure logs, traces and support correspondence can be retained for operational, security, support and legal purposes; their retention depends on the relevant service configuration and requirements. This policy does not promise immediate deletion or a fixed retention period for those records. Contact us to ask about information associated with a particular request.
Security and your choices
The endpoint uses HTTPS and exposes read-only tools for public data. No service can guarantee absolute security. You can stop sending requests or disconnect the connector in your client's settings at any time; this does not itself delete existing infrastructure logs or your client's conversation history.
For privacy questions, public-record corrections, or access, correction or deletion requests, email leo@pageonelab.com. Tell us which connector request or record is involved, using only the minimum information needed to locate it. We will respond in accordance with applicable law and may need to verify a request before acting on it.
Changes and contact
We will update this page and its effective date when these connector practices change. Contact SOTA2 at leo@pageonelab.com.