Skip to main content
These how-tos cover the three ways data moves in and out of Soundcheck with an import file (also called UEF): importing a file, exporting a workspace, and mapping a spreadsheet. Each works from the app or over MCP, so a person and an agent follow the same steps. MCP tool names changed when the catalog was merged. The old names still work: get_uef_schema is now get_import_format, validate_uef is check_import_file, and import_uef is import_file.

Import a file

1. Get the format

Call get_import_format on the public MCP endpoint (https://mcp.soundchecklive.io/public/mcp, no login). It returns the JSON Schema for the served schema_version, two example documents and a docs URL. The field reference lists the same fields for people.

2. Check the file

Call check_import_file with the file’s contents as an object or a JSON string (up to 256 KiB). It runs locally, needs no login and saves nothing. Fix every error it reports before you go on; see Read the check results. check_import_file can’t see your workspace, so a valid result doesn’t promise a clean import. The preview in the next step checks the file against your records, positions and permissions.

3. Preview against your workspace

On the member MCP endpoint (https://mcp.soundchecklive.io/mcp), call preview_import_file with the file. Nothing is written. It returns how many records would be new and how many would update existing ones, plus any errors.

4. Import

Call import_file with the same file. It writes the records into your workspace and may email organization invitations to team members in members[]. It is confirmation-gated, so your agent asks you before it writes, and it only runs on an organization you own or manage. Public MCP doesn’t have it. In the app, the same import is Settings → Data & knowledge → Import & export → Import a Soundcheck file. Over the REST API, it is POST /api/v1/organizations/{orgId}/imports/uef with the document as the body. Importing the same file again is safe: records match by external_id, then by natural key, so a second run updates instead of duplicating. See Identity and safe re-imports.

Export and restore a workspace

  1. Export. Call export_workspace on the member MCP endpoint, or click Download workspace on Settings → Data & knowledge → Import & export. MCP returns one import file, version 2.0 document with source.type: "EXPORT"; the app download is a zip with that document in uef/workspace.json plus a CSV per catalog.
  2. Restore. Pass the file to preview_import_file and then import_file on the destination organization, or upload the zip on Settings → Data & knowledge → Import & export.
An export is only what import can recreate: ledger lines, files, branding and integration secrets stay out, and money fields are withheld from callers who can’t view finances. Export a workspace and restore it lists exactly what comes back.

Map a CSV

A spreadsheet doesn’t need to be turned into an import file first. Its columns are mapped to the same field names, the same paths as in the field reference. In the app, drop the file on Settings → Data & knowledge → Import & export and follow Import your data and start running. Over MCP:
  1. create_import_from_paste with the CSV or TSV text and a target_type: EVENT, LEAD, INVENTORY or MEMBER. It creates an import job and parses the rows.
  2. get_import_field_catalog with the same target_type lists the fields a column can map to.
  3. map_import with the job ID. Leave out mapping to let Soundcheck map the columns itself, or pass a mapping object whose columns list each source column header with its uef_field path. For gigs, row_grouping: "group_by_external_id" inside mapping folds several rows with the same external ID into one gig.
  4. preview_import shows new and updated rows and errors without writing.
  5. commit_import writes the rows. It is confirmation-gated and can’t be undone from MCP.
Coming from another tool? The migration guides give the export steps and column mappings for Planning Center, Back On Stage, Master Tour, Stagera and LASSO.

Read the check results

check_import_file returns: Match on path and severity. The wording of message can change. The checks run in two passes. The first checks the shape: the version, known keys, types, required fields and formats. If that fails, it reports those errors and stops, so a second run can show more errors once you fix the first ones. For example, this gig has an end date before its start, a venue ref with no venue, and a lowercase currency code:

Field reference

Every field, with an example for each kind of record

Stability and versioning

What stays compatible within 2.x

Connect your AI (MCP)

Endpoints, sign-in and the full tool catalog

Coming from another system

CSV migration guides