Decide who can do what.On every door.
Set what each teammate can do and which assistants they can use. It applies in the app, the API, the CLI and their agents.
ActivityLogs
Mayagave Tom access to SupportApp
Tomadded Price list 2026MCP
Mayarevoked the key of OakmereApp
Priyaedited Sales guidelinesCLI
Lookio AIcreated the assistant Product docsApp
Mayaset a ceiling on Norvik, clientApp
Priyasynced help.acme.comAPI
Tomdeleted Old pricingApp
TTomMember
What he can do
Resources
ViewCreateEditDelete
Assistants
EditCreateDelete
Connections
API & CLIMCP
Assistants he can use
SupportProduct docsSales
One rule, every door
The app, the API, the CLI and MCP share one service layer. A permission refused in the app is refused everywhere.
Assistant by assistantChoose which assistants each member can use: a list, or all of them including future ones.
Agents includedA person’s CLI login and AI clients act with that person’s rights, and you can cap what they may hand over.
See what happened
Four logs: what people asked, what Lookio answered, what was indexed, and who changed what.
Where it fell shortRead widget and API conversations, and filter on thumbs down.
Who changed whatThe activity log shows each change and the door it came through, for a year.
Questions about control
Which plan has per-member permissions?
The Pro plan, with assigned assistants. On other plans, every member has every permission and sees every assistant. See pricing.
What can I set for each member?
Who can view, create, edit and delete resources; who can create, edit and delete assistants; who can use the API and the CLI, and MCP. And which assistants each person can use: a list, or all of them including future ones.
Do permissions also apply to the CLI and to AI agents?
Yes. A person who connects the CLI or an MCP client acts with their own permissions, and a right removed in the app is gone on their next call. An admin can also cap what members may hand to an AI client.
Can an admin read my chats?
No. A person’s chats in the app are personal: other members, admins and API keys do not see them. Widget, API key and MCP conversations are read by people who can edit assistants.
What do the logs keep?
Four logs: conversations, answers, indexing, and activity, which shows who changed what and through which door. Widget, API key and MCP conversations stay readable 7 days on Free, 90 days on Starter and a year on Pro after their last message. Activity is kept for a year.
Are my documents used to train models?
No. We do not use your content to train models, and the commercial APIs we use do not under their current terms. The detail is in the privacy policy.
