Skip to content
Intlexl

Security and governance

Control you can see.

Access, history and approvals are part of the product, so you can show exactly how your projects are governed.

Book a demoSee the agents

Access

Four layers, each one visible.

Organisation roles

Roles built from permission keys, managed by a separate access service.

Item access

Owner, editor or viewer on each workspace, sheet, dashboard, report and resource manager.

Administrator override

A permission for administrators, kept apart from everyday roles.

Organisation boundary

Access across organisations is denied.

History

A record of every change.

Who edited a cell, restructured a sheet or wrote through an MCP token is on the record. Agents keep their own history of runs, findings and tool calls.

  • Activity log

    Filter by person, item and action.

  • Row history

    See how a single task changed over time.

  • Agent history

    Findings, chases, reports, runs and tool calls, with retention you set.

AI under control

You set how far each agent may go.

Every agent runs at a tier you choose, and every output says whether a rule or a model wrote it.

  • Observe, Propose, Act

    Per agent, per organisation. Act is off unless an administrator enables it.

  • AI and Automated marks

    Counts and dates are always automated. Wording written by a model is labelled.

  • Approval before structure

    The assistant asks before it creates, deletes or restructures.

Automated

Forecast: Customer Portal is likely about 13 days late. Counts and dates come from your data.

AI wording

Priya, three tasks on Customer Portal are overdue. Could you update them today?

Delete the sheet Risk Register? This needs your approval.

Tokens

Tokens you scope, revoke and audit.

MCP clients connect with a personal token. Choose its scope, revoke it, refresh it, and see what it did.

FAQ

Questions people ask.

People belong to an organisation and hold roles made of permission keys, managed by a separate access service. On top of that, workspaces, sheets, dashboards, reports and resource managers each have owner, editor and viewer access set per person. Access across organisations is denied.

Yes. Cell edits, structural changes and writes made through MCP are recorded in an activity log with who, what and when. Agent runs and tool calls have their own history, with retention you control.

The assistant and MCP clients act through the signed-in person's permissions and cannot exceed them. Structural and destructive actions ask for approval first, and PM agents run at Observe, Propose or Act, per agent, with Act enabled only by an administrator.

A person creates a token with read, edit, structural or admin scope. Admin scope can only be created by an organisation administrator. Tokens can be revoked or refreshed at any time, and every call is audited.

Yes. In a demo we walk through roles, sharing, the activity log, MCP tokens and agent tiers, and answer your reviewers' questions directly.

Bring one plan. See it run.

In 30 minutes we set up a workspace like yours and show sheets, Gantt, dashboards, capacity and a PM agent at work.

Book a demoExplore the platform