Concept · Legibility
Make AI access
explicit.
Legibility means people and systems can see who can do what — grants, logs, and identity — instead of relying on tribal knowledge in side chats.
Idea
Hidden context does not scale.
When permissions live in someone's head, or in an unofficial bot token, organizations cannot govern AI. Legibility is the opposite: access you can name, revoke, and audit.
01
MCP grants you can revoke
Connect Cursor, Claude, and other clients through MCP client and server — with audited, revocable grants instead of permanent shadow access.
02
Audit logs on Enterprise
Security can review admin and agent activity, including MCP audit events, when they need evidence — not screenshots from chat.
03
SSO / SAML on Enterprise
Identity is the control plane. Enterprise wires FabricLoop into your IdP so AI and collab share the same access story.
In FabricLoop
Product mapping
SSO, audit logs, and admin console are standard Enterprise capabilities. When you need HIPAA, we'll walk through a BAA with you.
New. Better. Native. Fast.
Messaging with work attached. Connect Cursor, Claude, or ChatGPT via MCP to read and write notes, tasks, and news. Enterprise controls when you need them.
