Class AgentEngines (1.121.0)

AgentEngines(api_client_: google.genai._api_client.BaseApiClient)

API documentation for AgentEngines class.

Properties

memories

API documentation for memories property.

sandboxes

API documentation for sandboxes property.

sessions

API documentation for sessions property.

Methods

append_session_event

append_session_event(
    *,
    name: str,
    author: str,
    invocation_id: str,
    timestamp: datetime.datetime,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.AppendAgentEngineSessionEventConfig,
            vertexai._genai.types.AppendAgentEngineSessionEventConfigDict,
        ]
    ] = None
) -> vertexai._genai.types.AppendAgentEngineSessionEventResponse

Deprecated. Use agent_engines.sessions.events.append instead.

create

create(
    *,
    agent_engine: typing.Optional[typing.Any] = None,
    agent: typing.Optional[typing.Any] = None,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.AgentEngineConfig,
            vertexai._genai.types.AgentEngineConfigDict,
        ]
    ] = None
) -> vertexai._genai.types.AgentEngine

Creates an agent engine.

The Agent Engine will be an instance of the agent_engine that was passed in, running remotely on Vertex AI.

Sample src_dir contents (e.g. ./user_src_dir):

user_src_dir/
|-- main.py
|-- requirements.txt
|-- user_code/
|   |-- utils.py
|   |-- ...
|-- ...

To build an Agent Engine with the above files, run:

client = <xref uid="vertexai.Client">vertexai.Client</xref>(
    project="your-project",
    location="us-central1",
)
remote_agent = client.agent_engines.create(
    agent=local_agent,
    config=dict(
        requirements=[
            # I.e. the PyPI dependencies listed in requirements.txt
            "google-cloud-aiplatform[agent_engines,adk]",
            ...
        ],
        extra_packages=[
            "./user_src_dir/main.py", # a single file
            "./user_src_dir/user_code", # a directory
            ...
        ],
    ),
)

Parameters
Name	Description
`agent`	`Any` Optional. The Agent to be created. If not specified, this will correspond to a lightweight instance that cannot be queried (but can be updated to future instances that can be queried).
`agent_engine`	`Any` Optional. This is deprecated. Please use `agent` instead.
`config`	`AgentEngineConfig` Optional. The configurations to use for creating the Agent Engine.

Exceptions
Type	Description
`ValueError`	If the `project` was not set using `client.Client`.
`ValueError`	If the `location` was not set using `client.Client`.
`ValueError`	If `config.staging_bucket` was not set when `agent`
`i`	specified.:
`ValueError`	If `config.staging_bucket` does not start with "gs://".
`ValueError`	If `config.extra_packages` is specified but `agent`
`i`	None.:
`ValueError`	If `config.requirements` is specified but `agent` is None.
`ValueError`	If `config.env_vars` has a dictionary entry that does not
`correspon`	to an environment variable value or a SecretRef.:
`TypeError`	If `config.env_vars` is not a dictionary.
`FileNotFoundError`	If `config.extra_packages` includes a file or
`director`	that does not exist.:
`IOError`	If `config.requirements` is a string that corresponds to a
`nonexisten`	file.:

Returns
Type	Description
`AgentEngine`	The created Agent Engine instance.

create_memory

create_memory(
    *,
    name: str,
    fact: str,
    scope: dict[str, str],
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.AgentEngineMemoryConfig,
            vertexai._genai.types.AgentEngineMemoryConfigDict,
        ]
    ] = None
) -> vertexai._genai.types.AgentEngineMemoryOperation

Deprecated. Use agent_engines.memories.create instead.

create_session

create_session(
    *,
    name: str,
    user_id: str,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.CreateAgentEngineSessionConfig,
            vertexai._genai.types.CreateAgentEngineSessionConfigDict,
        ]
    ] = None
) -> vertexai._genai.types.AgentEngineSessionOperation

Deprecated. Use agent_engines.sessions.create instead.

delete

delete(
    *,
    name: str,
    force: typing.Optional[bool] = None,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.DeleteAgentEngineConfig,
            vertexai._genai.types.DeleteAgentEngineConfigDict,
        ]
    ] = None
) -> vertexai._genai.types.DeleteAgentEngineOperation

Delete an Agent Engine resource.

Parameters
Name	Description
`name`	`str` Required. The name of the Agent Engine to be deleted. Format: `projects/{project}/locations/{location}/reasoningEngines/{resource_id}` or `reasoningEngines/{resource_id}`.
`force`	`bool` Optional. If set to True, child resources will also be deleted. Otherwise, the request will fail with FAILED_PRECONDITION error when the Agent Engine has undeleted child resources. Defaults to False.
`config`	`DeleteAgentEngineConfig` Optional. Additional configurations for deleting the Agent Engine.

delete_memory

delete_memory(
    *,
    name: str,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.DeleteAgentEngineMemoryConfig,
            vertexai._genai.types.DeleteAgentEngineMemoryConfigDict,
        ]
    ] = None
) -> vertexai._genai.types.DeleteAgentEngineMemoryOperation

Deprecated. Use agent_engines.memories.delete instead.

delete_session

delete_session(
    *,
    name: str,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.DeleteAgentEngineSessionConfig,
            vertexai._genai.types.DeleteAgentEngineSessionConfigDict,
        ]
    ] = None
) -> vertexai._genai.types.DeleteAgentEngineSessionOperation

Deprecated. Use agent_engines.sessions.delete instead.

generate_memories

generate_memories(
    *,
    name: str,
    vertex_session_source: typing.Optional[
        typing.Union[
            vertexai._genai.types.GenerateMemoriesRequestVertexSessionSource,
            vertexai._genai.types.GenerateMemoriesRequestVertexSessionSourceDict,
        ]
    ] = None,
    direct_contents_source: typing.Optional[
        typing.Union[
            vertexai._genai.types.GenerateMemoriesRequestDirectContentsSource,
            vertexai._genai.types.GenerateMemoriesRequestDirectContentsSourceDict,
        ]
    ] = None,
    direct_memories_source: typing.Optional[
        typing.Union[
            vertexai._genai.types.GenerateMemoriesRequestDirectMemoriesSource,
            vertexai._genai.types.GenerateMemoriesRequestDirectMemoriesSourceDict,
        ]
    ] = None,
    scope: typing.Optional[dict[str, str]] = None,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.GenerateAgentEngineMemoriesConfig,
            vertexai._genai.types.GenerateAgentEngineMemoriesConfigDict,
        ]
    ] = None
) -> vertexai._genai.types.AgentEngineGenerateMemoriesOperation

Deprecated. Use agent_engines.memories.generate instead.

get

get(
    *,
    name: str,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.GetAgentEngineConfig,
            vertexai._genai.types.GetAgentEngineConfigDict,
        ]
    ] = None
) -> vertexai._genai.types.AgentEngine

Gets an agent engine.

Parameter
Name	Description
`name`	`str` Required. A fully-qualified resource name or ID such as "projects/123/locations/us-central1/reasoningEngines/456" or a shortened name such as "reasoningEngines/456".

get_memory

get_memory(
    *,
    name: str,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.GetAgentEngineMemoryConfig,
            vertexai._genai.types.GetAgentEngineMemoryConfigDict,
        ]
    ] = None
) -> vertexai._genai.types.Memory

Deprecated. Use agent_engines.memories.get instead.

get_session

get_session(
    *,
    name: str,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.GetAgentEngineSessionConfig,
            vertexai._genai.types.GetAgentEngineSessionConfigDict,
        ]
    ] = None
) -> vertexai._genai.types.Session

Deprecated. Use agent_engines.sessions.get instead.

list

list(
    *,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.ListAgentEngineConfig,
            vertexai._genai.types.ListAgentEngineConfigDict,
        ]
    ] = None
) -> typing.Iterator[vertexai._genai.types.AgentEngine]

List all instances of Agent Engine matching the filter.

Example Usage:

    import vertexai
client = <xref uid="vertexai.Client">vertexai.Client</xref>(project="my_project", location="us-central1")
for agent in client.agent_engines.list(
    config={"filter": "'display_name="My Custom Agent"'},
):
    print(agent.api_resource.name)

Parameter
Name	Description
`config`	`ListAgentEngineConfig` Optional. The config (e.g. filter) for the agents to be listed.

Returns
Type	Description
`Iterable[AgentEngine]`	An iterable of Agent Engines matching the filter.

list_memories

list_memories(
    *,
    name: str,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.ListAgentEngineMemoryConfig,
            vertexai._genai.types.ListAgentEngineMemoryConfigDict,
        ]
    ] = None
) -> typing.Iterator[vertexai._genai.types.Memory]

Deprecated. Use agent_engines.memories.list instead.

list_session_events

list_session_events(
    *,
    name: str,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.ListAgentEngineSessionEventsConfig,
            vertexai._genai.types.ListAgentEngineSessionEventsConfigDict,
        ]
    ] = None
) -> typing.Iterator[vertexai._genai.types.SessionEvent]

Deprecated. Use agent_engines.sessions.events.list instead.

list_sessions

list_sessions(
    *,
    name: str,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.ListAgentEngineSessionsConfig,
            vertexai._genai.types.ListAgentEngineSessionsConfigDict,
        ]
    ] = None
) -> typing.Iterator[vertexai._genai.types.Session]

Deprecated. Use agent_engines.sessions.list instead.

retrieve_memories

retrieve_memories(
    *,
    name: str,
    scope: dict[str, str],
    similarity_search_params: typing.Optional[
        typing.Union[
            vertexai._genai.types.RetrieveMemoriesRequestSimilaritySearchParams,
            vertexai._genai.types.RetrieveMemoriesRequestSimilaritySearchParamsDict,
        ]
    ] = None,
    simple_retrieval_params: typing.Optional[
        typing.Union[
            vertexai._genai.types.RetrieveMemoriesRequestSimpleRetrievalParams,
            vertexai._genai.types.RetrieveMemoriesRequestSimpleRetrievalParamsDict,
        ]
    ] = None,
    config: typing.Optional[
        typing.Union[
            vertexai._genai.types.RetrieveAgentEngineMemoriesConfig,
            vertexai._genai.types.RetrieveAgentEngineMemoriesConfigDict,
        ]
    ] = None
) -> typing.Iterator[vertexai._genai.types.RetrieveMemoriesResponseRetrievedMemory]

Deprecated. Use agent_engines.memories.retrieve instead.

update

update(
    *,
    name: str,
    agent: typing.Optional[typing.Any] = None,
    agent_engine: typing.Optional[typing.Any] = None,
    config: typing.Union[
        vertexai._genai.types.AgentEngineConfig,
        vertexai._genai.types.AgentEngineConfigDict,
    ]
) -> vertexai._genai.types.AgentEngine

Updates an existing Agent Engine.

This method updates the configuration of an existing Agent Engine running remotely, which is identified by its name.

Parameters
Name	Description
`name`	`str` Required. A fully-qualified resource name or ID such as "projects/123/locations/us-central1/reasoningEngines/456" or a shortened name such as "reasoningEngines/456".
`agent`	`Any` Optional. The instance to be used as the updated Agent Engine. If it is not specified, the existing instance will be used.
`agent_engine`	`Any` Optional. This is deprecated. Please use `agent` instead.
`config`	`AgentEngineConfig` Optional. The configurations to use for updating the Agent Engine.

Exceptions
Type	Description
`ValueError`	If the `project` was not set using `client.Client`.
`ValueError`	If the `location` was not set using `client.Client`.
`ValueError`	If `config.staging_bucket` was not set when `agent_engine`
`i`	specified.:
`ValueError`	If `config.staging_bucket` does not start with "gs://".
`ValueError`	If `config.extra_packages` is specified but `agent_engine`
`i`	None.:
`ValueError`	If `config.requirements` is specified but `agent_engine` is
`None.`
`ValueError`	If `config.env_vars` has a dictionary entry that does not
`correspon`	to an environment variable value or a SecretRef.:
`TypeError`	If `config.env_vars` is not a dictionary.
`FileNotFoundError`	If `config.extra_packages` includes a file or
`director`	that does not exist.:
`IOError`	If `config.requirements` is a string that corresponds to a
`nonexisten`	file.:

Returns
Type	Description
`AgentEngine`	The updated Agent Engine.