The SafeBase MCP Server is now Generally Available, giving your organization secure, permission-scoped AI access to your Trust Library, Trust Center, Accounts, Documents, and Questionnaires, directly from the AI tools your team already uses.
AI clients like Claude, ChatGPT, Cursor, and Claude Code are becoming part of how security and GRC teams get work done. The SafeBase MCP Server brings SafeBase into those workflows with an officially supported connection, so you can manage Trust Center content in plain language without leaving your AI tool of choice.
What's New
- Officially supported, GA availability: Connect SafeBase to Claude, ChatGPT, Cursor, Claude Code, and other MCP-compatible clients with full support.
- Expanded toolset: Now 70+ tools, including questionnaire answer editing and autofill retries, Trust Center content editing, content collections, tag management, translations, and document downloads.
- Admin-controlled, permission-scoped access: Org admins choose read-only or read + write access in Settings > Generative AI > MCP. Permissions are inherited from each member's existing SafeBase role, so AI actions stay limited to what that person can already do in the SafeBase UI.
Learn more at: SafeBase MCP Server Guide