Manage data
Everything about the data itself - tables, domains, storage connections, and workspaces.
At the centre is the table - a real Delta Lake table at a prefix in your own object storage; Eddytor connects to it, never copies it. Every write is a numbered version you can diff, time-travel, and roll back, and the data stays readable by any Delta-compatible tool.
Fully-qualified names
Every table is identified by a three-part fully-qualified name (FQN):
catalog.schema.table - e.g. eddytor.sales.orders, the sample used
throughout these docs.
| Part | Meaning |
|---|---|
catalog | Top-level grouping. On self-host this is always eddytor. |
schema | A namespace within the catalog (e.g. sales, raw, curated). |
table | The table name. |
The CLI takes the parts as separate arguments (eddytor get history eddytor sales orders), REST as path segments, MCP as one dotted string
({ "table": "eddytor.sales.orders" }). MCP tools often show the table
component UUID-prefixed (e.g. eddytor.cfg_abc123.a1b2…_demo_products) -
the prefix is the table's stable UUID, so
renaming or moving doesn't change identity.
Get the exact string from list_tables. Quoting FQNs in raw SQL is covered on
the query page.
Before you start
- A registered storage connection - tables live in
your storage, so nothing can be created until a storage config exists. Capture
its
config_id(eddytor get storage/list_storage_configs). - Write access - creating tables and managing schemas requires the Builder role (or Admin); Editor covers row changes, Viewer covers reads. See Roles & scopes.
- Know your
location- creating a table takes a full storage URL under a registered connection, e.g.s3://your-bucket/master-data. (The CLI/REST infer it from the storage config + path; the MCPcreate_tabletakes the full URL.)
In this section
- Connect storage - register the S3 / GCS / Azure Blob / Fabric storage your tables live in.
- Create & model - create tables and folders, column types, schema evolution, primary keys.
- Load & edit - import, insert, merge, delete, validate.
- Domains - lock columns to governed value sets.
- Read & explore - query, SQL, profile, AI analysis.
- Operate - history, rollback and restore, rename, optimize and vacuum, drop.
- Security & audit - row and column policies, the audit log.
- Workspaces - containers inside an organisation that scope storage sharing, credentials, and what a session can see.