Work with your team

A workspace is shared. This page covers who gets in, what they can reach once they’re there, and how to see what they did.

Two different meanings of “person”

Worth separating up front, because they’re easy to conflate:

A person in Configuration → People is somebody your models describe. They have a role, a level, a location, and they can be placed into positions. They have no login.
A member of your organization is somebody who uses Resorcerer. They sign in, build models, and change things.

The same human is often both, but the records are independent. Adding someone to People does not invite them; inviting someone does not put them in your org chart.

Inviting members

Invite people from the organization switcher at the bottom of the sidebar, under Organization settings → Members. They get an email, and once they accept they land in your workspace.

Members are either administrators, who can change organization-wide settings, manage members, and reach every model including private ones, or ordinary members, who can work with the workspace’s models and configuration.

Somebody can belong to several organizations; the switcher moves them between the ones they’re in, and nothing is shared across the boundary.

Every page in Resorcerer lives under its organization’s address. When an organization has a Slug, links use it, so a link you share reads like /org/example-co/models rather than a long string of letters and numbers. Older links that still carry the long form keep working and switch to the slug when opened.

An administrator sets the slug on the Organizations page, in the Slug field of the Create organization or Edit organization dialog. A slug uses lowercase letters, numbers and hyphens. When your organization’s sign-in is managed by your identity provider, the slug comes from there and can’t be edited in Resorcerer.

Restricting a model

By default every model is visible to everyone in the workspace. Marking one private limits it to administrators and the people you name.

Open Manage access on the model and add members individually:

Level	Can
Viewer	See the model, but change nothing.
Editor	See and edit it, but not change who else has access.
Owner	Everything, including managing access.

Administrators always have owner-level access to private models, so you can’t lock yourself out of your own workspace.

Use it for a restructure that isn’t ready to be seen, or for planning that names individuals.

Locking a model against edits

Access control decides who; protection decides how. A protected model refuses direct edits from everyone, regardless of their access level, and changes only through a reviewed merge.

The two work together: the model of your live organization is typically visible to everyone and protected against everyone.

Discussing a model

Comments let you talk about a model where the question is. Leave one on the model as a whole, on a unit, or on a position, and anyone who can see the model can reply in the thread, including on protected models, since a comment changes nothing in the model itself.

Units and positions with open comments carry an amber comment badge in the Diagram, Org Chart, Units and Slots views; click it to read the thread and reply. To start a thread on a unit, hover it and the badge appears faintly. A position’s first thread starts from the Comments tab of its details panel. The Show comments toggle in the Diagram and Org Chart toolbars hides the badges when you want a clean picture.

While a model has open problems or open comment threads, the model header shows a count of each; click a count to go to its view.

The Comments view lists every open thread, grouped by what it is about, with a link to the unit or position. When a thread is settled, Resolve it: it leaves the badges and the list, and stays available on the Comments view behind the Show resolved toggle, where it can be reopened. Authors can edit or delete their own comments; administrators can delete anyone’s.

Seeing what changed

Every model has an Audit Log view — what changed, when, and who did it — and there’s an organization-wide log covering everything else in the workspace. A change made through an access token says so next to the person: an AI agent’s change carries an MCP chip with the name of the agent connection, and a set of changes the agent wrote back in one go also shows its batch and which operation in the batch made it.

Deletions are recoverable. Deleted models, units, positions, and people go to Trash, where they can be restored until they’re purged.

Automation and integrations

Members aren’t the only things that reach a workspace. Access tokens let scripts and agents in, scoped to what they need, and are created under My Settings → Access Tokens; administrators can review every token in the organization from the Access Tokens page. Registered apps let an internal tool or a partner’s app sign your members in, with permissions an administrator chooses under Organization Settings → Registered Apps. Outgoing webhooks notify another system when something changes.

See Connect AI agents, Register an app and APIs.

Models — private and protected models.
Compare and merge — the reviewed way into a protected model.
APIs — programmatic access.