Error Faced: The format for the tool name must be: \"httpVerb__actionGroupName__apiName\".}]","role":"user"}

I am getting repeated tool format name issue despite it says a tool of same type is available while executing agent with added action group.

I have added action group using 'Define with function details' and 'Define with API schemas' none of them seems to work, using 'Define with API schemas' got me closer to what agent is expecting tool name so I have attached it's traces but I am still not getting reponse and it is not able to trigger lambda function too.

The available action is GET__customer-support-agent__getCustomerId. (6) This action requires the email parameter, which is already provided by the User. (7) I have everything I need to proceed.}]","role":"assistant"},{"content":"[{text=The tool name format is incorrect. The format for the tool name must be: "httpVerb__actionGroupName__apiName".}]","role":"user"}

The available action is GET__customer-support-agent__getCustomerId. (6) This action requires the email parameter, which is already provided by the User. (7) I have everything I need to proceed.}]","role":"assistant"},{"content":"[{text=The tool name format is incorrect. The format for the tool name must be: "httpVerb__actionGroupName__apiName".}]","role":"user"}

Model used :- amazon.nova-pro-v1:0

Lambda Function (simple returns a random value, no deep logic) used attached below :-

import json, random

def unpack_parameters(parameters):
    result = {}
    for param in parameters:
        name = param.get("name")
        value = param.get("value")
        
        if name is not None:
            if value is not None:
                # Try to convert to int or float if possible
                try:
                    result[name] = int(value)
                except ValueError:
                    try:
                        result[name] = float(value)
                    except ValueError:
                        result[name] = value  # Keep as string if not a number
            else:
                result[name] = None  # Handle missing value
    
    return result

def lambda_handler(event, context):
    
    agent = event['agent']
    actionGroup = event['actionGroup']
    function = event['function']
    parameters = event.get('parameters', [])

    unpacked = unpack_parameters(parameters)

    responseBody =  {
        "TEXT": {
            "body": ""
        }
    }

    if function == "getCustomerId":
        email = unpacked.get("email", None)
        name = unpacked.get("name", None)
        phone = unpacked.get("phone", None)

        if any(value is not None and value != "" for value in (email, name, phone)):
            customer_id = random.randint(1000, 9999)
            responseBody['TEXT']['body'] = f"{{'id':{customer_id}}}"
        else:
            responseBody['TEXT']['body'] = "{'error':'Customer not found.'}"

    elif function == "sendToSupport":
        custId = unpacked.get("custId", None)
        supportSummary = unpacked.get("supportSummary", None)
        
        if all(value is not None and value != "" for value in (custId, supportSummary)):
            support_id = random.randint(1000, 9999)
            responseBody['TEXT']['body'] = f"{{'supportId': {support_id}}}"
        else:
            responseBody['TEXT']['body'] = "{'error':'Details missing.'}"

    else:
        responseBody['TEXT']['body'] = "{'error':'unknown function'}"


    action_response = {
        'actionGroup': actionGroup,
        'function': function,
        'functionResponse': {
            'responseBody': responseBody
        }

    }

    dummy_function_response = {'response': action_response, 'messageVersion': event['messageVersion']}
    print("Response: {}".format(dummy_function_response))

    return dummy_function_response