jq : update json content when path is variable

I have the following dataset in a file :

{
   "storageSummary": 6.8624392E13,
   "audit1": {
         "auditScore": 1,
         "suspensionScore": 1,
         "onlineScore": 0.99743587,
         "satelliteName": "us2.storj.io:7777"
      },
   "audit2": {
         "auditScore": 1,
         "suspensionScore": 1,
         "onlineScore": 0.9992917,
         "satelliteName": "saltlake.tardigrade.io:7777"
      },
   "audit3": {
         "auditScore": 1,
         "suspensionScore": 1,
         "onlineScore": 0.99930555,
         "satelliteName": "ap1.storj.io:7777"
      }
}

I want to update the content (I'm storing the json into a variable, since I make several updates, then I'll create a new file), but the path of the data to update is variable :

For example, I have :

DS_SRC_FILE="test.json"

DS_OBJECT=$(jq -c '' ${DS_SRC_FILE})

ds="audit1"
new_name="toto"

jq --arg ds ${ds} --arg name ${new_name} '.[]|select(.value==$ds)|=(.satellitename = $name)' <<< ${DS_OBJECT}

So, of course, this is failing with parse error: Invalid numeric literal at line 2, column 36

What's the right way to do that please ?

JSON doesn't use single quotes. Use double quotes instead, or leave the number unquoted.

The command will still fail: starting with .[] will throw away the keys of the object and return the values, so you don't have access to the "audit1" anymore.

Also, to change the value, you can use the plain = (and without select) as the right hand side is not sensitive to context:

jq --arg ds "$ds" --arg name "$new_name" '.[$ds]["satelliteName"] = $name' <<< "$DS_OBJECT"

Also note that I double quoted the shell variables which should make the command work even if they contained whitespace.