classify_inbox_items

Correct sentiment, language, intent, needsResponse, spam or urgent on inbox items. Provide items or a structured filter and at least one field. Null clears a field. Pins the correction against automatic classification. Setting urgent true (safety, harassment or legal matters) also sets critical priority and needsResponse and alerts the client's team once. Checks all selected clients before writes; never invokes the model or sends a message.

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "classify_inbox_items",
    "arguments": {}
  }
}

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