steamship.agents.examples package#

Submodules#

steamship.agents.examples.chatgpt_api module#

Minimum viable AgentService implementation.

This will result in an agent that effectively acts like ChatGPT.

class steamship.agents.examples.chatgpt_api.MyAssistant(**kwargs)[source]#

Bases: AgentService

steamship.agents.examples.chatgpt_with_web_and_telegram module#

class steamship.agents.examples.chatgpt_with_web_and_telegram.ChatGpt(**kwargs)[source]#

Bases: AgentService

Minimal implementation of an Agent that just chats with you via the web.

steamship.agents.examples.document_qa_agent module#

class steamship.agents.examples.document_qa_agent.ExampleDocumentQAService(**kwargs)[source]#

Bases: AgentService

DocumentQAService is an example AgentService that exposes: # noqa: RST201

  • A few authenticated endpoints for learning PDF and YouTube documents:

    /index_url

    { url }

    /index_text { text }

  • An unauthenticated endpoint for answering questions about what it has learned

This agent provides a starter project for special purpose QA agents that can answer questions about documents you provide.

USED_MIXIN_CLASSES: List[Type[PackageMixin]] = [<class 'steamship.invocable.mixins.indexer_pipeline_mixin.IndexerPipelineMixin'>, <class 'steamship.invocable.mixins.file_importer_mixin.FileImporterMixin'>, <class 'steamship.invocable.mixins.blockifier_mixin.BlockifierMixin'>, <class 'steamship.invocable.mixins.indexer_mixin.IndexerMixin'>]#

steamship.agents.examples.example_assistant module#

class steamship.agents.examples.example_assistant.MyFunctionsBasedAssistant(**kwargs)[source]#

Bases: AgentService

MyFunctionsBasedAssistant is an example AgentService that exposes a single test endpoint for trying out Agent-based invocations. It is configured with two simple Tools to provide an overview of the types of tasks it can accomplish (here, search and image generation).

class AgentConfig(*, modelName: str = 'gpt-4')[source]#

Bases: Config

model_name: str#
classmethod config_cls() → Type[Config][source]#

Returns the configuration object for the Invocable.

By default, Steamship packages and plugins will not take any configuration. Steamship packages and plugins may declare a configuration object which extends from Config, if needed, as follows:

class MyPackageOrPlugin:
class MyConfig(Config):

…

@classmethod def config_cls(cls):

return MyPackageOrPlugin.MyConfig

steamship.agents.examples.example_assistant_with_caching module#

class steamship.agents.examples.example_assistant_with_caching.MyCachingAssistant(**kwargs)[source]#

Bases: AgentService

MyCachingAssistant is an example AgentService that exposes a single test endpoint for trying out Agent-based invocations. It is configured with two simple Tools to provide an overview of the types of tasks it can accomplish (here, search and image generation).

get_max_actions_per_run() → int[source]#

Save the max_actions_per_run value so that it will be reloaded upon next request.

set_max_actions_per_run(value: int)[source]#

Save the max_actions_per_run value so that it will be reloaded upon next request.

steamship.agents.examples.example_react_assistant module#

class steamship.agents.examples.example_react_assistant.MyAssistant(**kwargs)[source]#

Bases: AgentService

MyAssistant is an example AgentService that exposes a single test endpoint for trying out Agent-based invocations. It is configured with two simple Tools to provide an overview of the types of tasks it can accomplish (here, search and image generation).

steamship.agents.examples.fact_learner module#

class steamship.agents.examples.fact_learner.FactLearner(**kwargs)[source]#

Bases: AgentService

FactLearner is an example AgentService contains an Agent which:

  1. Learns facts to a vector store

  2. Can answer questions based on those facts

steamship.agents.examples.telegram_bot module#

class steamship.agents.examples.telegram_bot.TelegramBot(**kwargs)[source]#

Bases: AgentService

Deployable Multimodal Agent that lets you talk to Google Search & Google Images.

NOTE: To extend and deploy this agent, copy and paste the code into api.py.

class TelegramBotConfig(*, telegramBotToken: str)[source]#

Bases: Config

telegram_bot_token: str#
config: TelegramBotConfig#
classmethod config_cls() → Type[Config][source]#

Returns the configuration object for the Invocable.

By default, Steamship packages and plugins will not take any configuration. Steamship packages and plugins may declare a configuration object which extends from Config, if needed, as follows:

class MyPackageOrPlugin:
class MyConfig(Config):

…

@classmethod def config_cls(cls):

return MyPackageOrPlugin.MyConfig

Module contents#