Getting started

Introduction

Basemind is an open workspace where people and AI agents build together. It combines a no-code database, an application builder, dashboards and automations with a project memory that agents can read and write through the Model Context Protocol (MCP).

Basemind is based on the open-source Baserow platform, so everything you know about spreadsheet-style databases carries over.

What you can build

Building block What it is for
Databases Tables with typed fields, relations, formulas and views. The source of truth for your data.
Application builder Internal tools, client portals and websites on top of your tables, published on your own domain.
Dashboards Summary numbers and charts that stay in sync with your tables.
Automations Workflows that react to events or schedules and act on your data or external services.
Projects A folder for one initiative: its databases, memory, documents, reports and agent activity.
Reports Read-only pages with metrics, charts and tables that you can publish with a public link.

How the pieces fit together

Workspace
├── Databases ── Tables ── Fields / Views / Rows
├── Applications (builder) ── Pages ── Elements / Data sources
├── Dashboards ── Widgets
├── Automations ── Workflows ── Nodes
└── Projects
    ├── attached databases, apps and dashboards
    ├── Memory (facts, decisions, assumptions…)
    ├── Documents
    ├── Reports ── published at /{handle}/{project}/{report}
    └── Agent activity (audit log, approvals, undo)

A workspace is the unit of collaboration and permissions. Everything else lives inside one.

Humans and agents, same rules

Every change an agent makes goes through the same permission checks, undo history and realtime updates as a change made in the UI. On top of that, each project has an autonomy level that decides whether agent changes need your approval first.

Where to go next