Create and run a workspace — settings, team, agents, context, and skills.
A workspace is the home for your team, agents, motions, and the context Monnet uses to help you decide. This page covers the workspace-side tasks: creating it, configuring settings, managing people and agents, giving Monnet persistent context, and managing skills.For the concept itself — what a workspace is, roles, how it fits in the Monnet model — see How Monnet works.
From the sidebar, click New workspace and give it a name. The URL slug is generated from the name and can’t be changed after creation — your workspace lives at monnet.ai/your-slug. Once created, you’re the first member with the Owner role.
Open the workspace’s settings (the edit affordance next to its name in the sidebar). Settings are organized into five tabs: General, Team, Context, Skills, and Integrations.
Click Invite a member, enter an email, and pick a workspace role:
Owner — can do anything, including delete the workspace.
Admin — can manage the team and settings, but can’t delete the workspace.
Member — collaborates on motions.
Monnet emails the invitee a link to accept or decline. You can resend or cancel a pending invitation from the team list.Click a member to open their panel, where admins and owners can set:
Function — what this person does on the team (e.g. “Designer”).
When to involve this person — optional guidance telling Monnet when to loop them in, and when not to.
Role — switch between Owner, Admin, and Member.
From the same panel you can Remove from workspace (or Leave workspace for yourself).
A workspace must keep at least one Owner — you can’t remove or downgrade the last one.
Agents are AI teammates that execute motion plan steps. In the AI Agents sub-view, click Onboard a new agent and follow the wizard. The available connector is Local Runner (Claude Code), which runs Claude Code on a machine you control.Once onboarded, an agent appears in the team list and can be assigned to plan steps. For the full setup — provisioning a runner, the per-OS install command, and authorizing the device — see Agents.
Your workspace’s policy file — guidelines and policies Monnet uses when drafting and discussing motions. Edit it as plain markdown. This is the persistent brief a new teammate (human or agent) would need to be productive.See How Monnet works — MONNET.md for what to put in it.
A running digest that Monnet auto-maintains when motions close — it accumulates what the workspace has decided and learned over time. It’s editable: admins can prune or correct it, but you don’t have to write it yourself. A fresh workspace shows it empty until the first motion closes.
Upload supporting documents (handbook, style guide, product requirements) that Monnet can reference across motions. Use Add file to upload; remove a file from the same list. These are workspace-wide knowledge, distinct from files attached to a single motion.Click Save changes to persist edits to MONNET.md and MEMORY.md.
The Skills tab holds reusable instructions for specific scenarios — a competitive-brief template, an incident report format, an onboarding checklist. Click Add skill and fill in a Name, a short Description, and the long-form Content (instructions). Open an existing skill to edit it or delete it. Skills are workspace-scoped and available to Monnet on every motion.
The Integrations tab connects external tools. Today that’s Slack — see the Slack integration page for connecting, importing members, and how teammates take part in motions from Slack DMs.
