MCPOutils MCP/Emplacements/bulk_update_locations

bulk_update_locations

Writelocations:write

Apply the same profile patch, publisher overrides, and/or tag add/remove to many locations at once, optionally publishing each to Google afterward. Useful for changes like "add the tag 'holiday-hours-2026' to these 40 locations" or "turn off online ordering everywhere in this client". Capped at 100 locations per call — extra ids beyond that are silently skipped (check the returned `skipped` count). Tags are merged per-location (added/removed against each location's existing tags), never wholesale replaced. Returns a per-location result so partial failures are visible.

Paramètres

locationIdsstring[]obligatoire

The locations to update (required, up to 100).

patchobjectfacultatif

Base profile fields to set on every location, same shape as update_location's fields (e.g. { description, phone, tags }), including hours (regularHours, moreHours, specialHours — e.g. to close every location in a client on the same date) and services. Omit fields you don't want to change. NOTE: `attributes` here must be the stored array form `[{ id, value }]`, NOT update_location's { id: value } map — this patch is passed through as given. For a change to ONE location prefer update_location, which validates and converts each field.

overridesobjectfacultatif

Per-publisher overrides to apply to every location, keyed by publisher (google/facebook/apple/bing), each an object of profile fields.

addTagsstring[]facultatif

Internal tags to add to every location (merged with each location's existing tags).

removeTagsstring[]facultatif

Internal tags to remove from every location.

publishbooleanfacultatif

If true, push each updated location to Google immediately afterward.

Exemple de requête

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "bulk_update_locations",
    "arguments": {
      "locationIds": [
        "string"
      ]
    }
  }
}

Exemple de réponse

{
  "jsonrpc": "2.0",
  "id": 1,
  "result": {
    "content": [
      {
        "type": "text",
        "text": "..."
      }
    ]
  }
}

La forme de la réponse est indicative — le contenu réel dépend de l'outil et de vos données.