Public-source policy
Privacy rules
Deep Research Archives is designed for public research material—not personal records, customer datasets, confidential operations, or secrets.
Last updated: July 17, 2026.
Account and public contribution data
Account creation stores a username and a password hash. When you publish, the website associates the public post with your username. Do not use a username or submission text that reveals private information you do not want displayed publicly.
Material that must not be submitted
- Personal contact, identity, health, financial, or location data.
- Customer, client, account, tenant, or non-public user information.
- Passwords, API keys, access tokens, private keys, or authentication headers.
- Internal endpoints and private session, request, trace, or incident identifiers.
- Confidential documents or material you lack permission to publish.
Automated checks can detect common patterns, but they cannot understand every context. Contributors must perform a complete manual review.
MCP data boundary
The public MCP tools are anonymous and read-only. They return the archive item ID, title, public text, archive URL, original public source URL, source type, citation, and archive date. They omit username, account ID, password, session, and rate-limit data. A public record that triggers the sensitive-data screen is not returned through MCP.
Rate limiting and infrastructure
The application uses a request IP address transiently as a rate-limit key for sign-in and sign-up protection and does not intentionally print that value in application logs. Hosting, database, network, or security providers may process request metadata under their own policies and legal obligations.
Questions and removal requests
To report sensitive material or request review, contact [email protected] and include the public archive URL. Do not email the sensitive value itself. You can also read the submission terms.