edisyl data-sources & connections

data-sources

Creating a data source via the CLI auto-creates a pack from the source name (pass --pack-id to attach to an existing pack instead; that needs contribute access to it).

edisyl data-sources test-connection --type <connector> --config-file <f>
edisyl data-sources create --name "<name>" --type <connector> --config-file <f> --database <db>
edisyl data-sources introspect <id>     # re-run table discovery only
edisyl data-sources enrich <id>         # re-study every imported model (after schema changes)
edisyl data-sources models <id>         # list indexed models (the names `study` takes)
edisyl data-sources list

Connectors: edisyl data-sources create --help lists the types available.

Flag	Behavior
--config-file <path> / --config-json '<inline>'	Connector-specific JSON. Exactly one, never both. Shape is validated server-side by the connector: a wrong key surfaces as a connector error, not a flag error.
--pack-id <id>	Attach to an existing pack instead of auto-creating one. Requires contribute access.
--selection-rule <scope>=<identifier>	Narrow the import. scope is table, schema, or database. Repeat per rule. Both = and : work as separator. Accepted only on create; neither import nor enrich takes it, so a rule can’t be widened without recreating the source.
--connection-id <id>	Back this data source with an existing reusable connection instead of fresh credentials.
--org-id <uuid>	Override the org (defaults to the authenticated profile’s org). Not registered on data-sources list (org-scoped by the auth token) or test-connection (validates an unsaved config, tied to no org).

test-connection returns { success: true, durationMs } on success, or { success: false, message } on failure: the message names the actual problem.

create fires two things atomically: introspect (table list) and an import that, by default, also studies every discovered model. That study is LLM work, and its results land as learning candidates that a separate review step applies a little later (minutes to an hour), after which the source is answerable. enrich re-runs that study over every model; use it after a schema change, not as part of initial setup. Tables added to the warehouse later are not catalogued automatically, see Adding tables to a connected source.

stratum

edisyl stratum status -d <data-source-id>

Watches the enrichment pipeline behind a data source: what’s imported, what’s modeled, and how fresh the semantic layer is.

connections

A connection holds provider credentials for the org, separate from any one data source. SQL providers back data sources (data-sources create --connection-id); API providers attach straight to a pack.

edisyl connections list
edisyl connections create --provider <provider> --name "<name>" --pack-id <id> --token-stdin
edisyl connections databases <connection-id>    # what a SQL connection can see

Use --token-stdin (pipe the token in) so it never lands in shell history or a process listing:

printf '%s' "$LINEAR_API_KEY" | edisyl connections create \
  --provider linear --name "Tracker" --pack-id <pack-id> --token-stdin