Given the following XML in a table column:

<UserFields>
  <UserField>
    <Name>UserName</Name>
    <Value>test_user</Value>
  </UserField>
  <UserField>
    <Name>Age</Name>
    <Value>24</Value>
  </UserField>
  <UserField>
    <Name>Password</Name>
    <Value>test1234!</Value>
  </UserField>
  <UserField>
    <Name>AgentPassword</Name>
    <Value>1234Test!</Value>
  </UserField>
</UserFields>

I want to query the table and update the XML to replace all <Value> nodes for any node that contains some form of ‘password’. So in the given example, I would want to replace the <Value> values of the last two <UserField> nodes to wind up with:

<UserFields>
  <UserField>
    <Name>UserName</Name>
    <Value>test_user</Value>
  </UserField>
  <UserField>
    <Name>Age</Name>
    <Value>24</Value>
  </UserField>
  <UserField>
    <Name>Password</Name>
    <Value>********</Value>
  </UserField>
  <UserField>
    <Name>AgentPassword</Name>
    <Value>********</Value>
  </UserField>
</UserFields>

I was able to come up with this:

SET @MyDoc.modify('
  replace value of (//UserField[Name="Password"]/Value/text())[1]
  with "********"
');

But it requires a hardcoded value of the Name node — it will find and replace the Value of the node with Password value but not the Value node with value of AgentPassword. Essentially, I’m trying to say [Name contains "Password"] instead of =.