0.31.0

Added

• Parameter meta: dict on BaseEvent.

Changed

• BREAKING: Drivers, Loaders, and Engines now raise exceptions rather than returning ErrorArtifacts.
• BREAKING: Parameter driver on BaseConversationMemory renamed to conversation_memory_driver.
• BREAKING: BaseConversationMemory.add_to_prompt_stack now takes a prompt_driver parameter.
• BREAKING: BaseConversationMemoryDriver.load now returns tuple[list[Run], dict]. This represents the runs and metadata.
• BREAKING: BaseConversationMemoryDriver.store now takes runs: list[Run] and metadata: dict as input.
• BREAKING: Parameter file_path on LocalConversationMemoryDriver renamed to persist_file and is now type Optional[str].
• Defaults.drivers_config.conversation_memory_driver now defaults to LocalConversationMemoryDriver instead of None.
• CsvRowArtifact.to_text() now includes the header.

Fixed

• Parsing streaming response with some OpenAI compatible services.
• Issue in PromptSummaryEngine if there are no artifacts during recursive summarization.
• Issue in GooglePromptDriver using Tools with no schema.
• Missing maxTokens inference parameter in AmazonBedrockPromptDriver.
• Incorrect model in OpenAiDriverConfig's text_to_speech_driver.
• Crash when using CohereRerankDriver with CsvRowArtifacts.

0.30.2

Fixed

• Ensure thread safety when publishing events by adding a thread lock to batch operations in BaseEventListenerDriver.
• FileManagerTool failing to save Artifacts created by ExtractionTool with a CsvExtractionEngine.

0.30.1

Fixed

• CsvExtractionEngine not using provided Rulesets.
• Docs examples for Extraction Engines not properly passing in schemas.

0.30.0

Added

• AstraDbVectorStoreDriver to support DataStax Astra DB as a vector store.
• Ability to set custom schema properties on Tool Activities via extra_schema_properties.
• Parameter structure to BaseTask.
• Method try_find_task to Structure.
• TranslateQueryRagModule RagEngine module for translating input queries.
• Global event bus, griptape.events.EventBus, for publishing and subscribing to events.
• Global object, griptape.configs.Defaults, for setting default values throughout the framework.
• Unique name generation for all RagEngine modules.
• ExtractionTool for having the LLM extract structured data from text.
• PromptSummaryTool for having the LLM summarize text.
• QueryTool for having the LLM query text.
• Support for bitshift composition in BaseTask for adding parent/child tasks.
• JsonArtifact for handling de/seralization of values.
• Chat.logger_level for setting what the Chat utility sets the logger level to.
• FuturesExecutorMixin to DRY up and optimize concurrent code across multiple classes.
• utils.execute_futures_list_dict for executing a dict of lists of futures.
• GriptapeCloudConversationMemoryDriver to store conversation history in Griptape Cloud.
• griptape.utils.decorators.lazy_property for creating lazy properties.

Changed

• BREAKING: Removed all uses of EventPublisherMixin in favor of EventBus.
• BREAKING: Removed EventPublisherMixin.
• BREAKING: Removed Pipeline.prompt_driver and Workflow.prompt_driver. Set this via griptape.configs.Defaults.drivers.prompt_driver instead. Agent.prompt_driver has not been removed.
• BREAKING: Removed Pipeline.stream and Workflow.stream. Set this via griptape.configs.Defaults.drivers.prompt_driver.stream instead. Agent.stream has not been removed.
• BREAKING: Removed Structure.embedding_driver, set this via griptape.configs.Defaults.drivers.embedding_driver instead.
• BREAKING: Removed Structure.custom_logger and Structure.logger_level, set these via logging.getLogger(griptape.configs.Defaults.logger_name) instead.
• BREAKING: Removed BaseStructureConfig.merge_config.
• BREAKING: Renamed StructureConfig to DriversConfig, moved to griptape.configs.drivers and renamed fields accordingly.
• BREAKING: RagContext.output was changed to RagContext.outputs to support multiple outputs. All relevant RAG modules were adjusted accordingly.
• BREAKING: Removed before and after response modules from ResponseRagStage.
• BREAKING: Moved ruleset and metadata ingestion from standalone modules to PromptResponseRagModule.
• BREAKING: Dropped Client from all Tool names for better naming consistency.
• BREAKING: Dropped _client suffix from all Tool packages.
• BREAKING: Added Tool suffix to all Tool names for better naming consistency.
• BREAKING: Removed TextArtifactStorage.query and TextArtifactStorage.summarize.
• BREAKING: Removed TextArtifactStorage.rag_engine, and TextArtifactStorage.retrieval_rag_module_name.
• BREAKING: Removed TextArtifactStorage.summary_engine, TextArtifactStorage.csv_extraction_engine, and TextArtifactStorage.json_extraction_engine.
• BREAKING: Removed TaskMemory.summarize_namespace and TaskMemory.query_namespace.
• BREAKING: Removed Structure.rag_engine.
• BREAKING: Split JsonExtractionEngine.template_generator into JsonExtractionEngine.system_template_generator and JsonExtractionEngine.user_template_generator.
• BREAKING: Split CsvExtractionEngine.template_generator into CsvExtractionEngine.system_template_generator and CsvExtractionEngine.user_template_generator.
• BREAKING: Changed JsonExtractionEngine.template_schema from a run argument to a class attribute.
• BREAKING: Changed CsvExtractionEngine.column_names from a run argument to a class attribute.
• BREAKING: Removed JsonExtractionTask, and CsvExtractionTask use ExtractionTask instead.
• BREAKING: Removed TaskMemoryClient, use QueryClient, ExtractionTool, or PromptSummaryTool instead.
• BREAKING: BaseTask.add_parent/child now take a BaseTask instead of str | BaseTask.
• Engines that previously required Drivers now pull from griptape.configs.Defaults.drivers_config by default.
• BaseTask.add_parent/child will now call self.structure.add_task if possible.
• BaseTask.add_parent/child now returns self, allowing for chaining.
• Chat now sets the griptape logger level to logging.ERROR, suppressing all logs except for errors.

Fixed

• JsonExtractionEngine failing to parse json when the LLM outputs more than just the json.
• Exception when adding ErrorArtifact's to the Prompt Stack.
• Concurrency bug in BaseVectorStoreDriver.upsert_text_artifacts.
• Schema issues with Tools that use lists.
• Issue with native Tool calling and streaming with GooglePromptDriver.
• Description not being used properly in StructureRunTool.

0.29.2

Fixed

• Workflow threads not being properly cleaned up after completion.
• Crash when ToolActions were missing output due to an ActionsSubtask exception.

0.29.1

Changed

• Remove BaseTextArtifact, revert CsvRowArtifact to subclass TextArtifact.

Fixed

• Missing extra for drivers-text-to-speech-elevenlabs.

0.29.0

Added

• Native function calling support to OpenAiChatPromptDriver, AzureOpenAiChatPromptDriver, AnthropicPromptDriver, AmazonBedrockPromptDriver, GooglePromptDriver, OllamaPromptDriver, and CoherePromptDriver.
• OllamaEmbeddingDriver for generating embeddings with Ollama.
• GriptapeCloudKnowledgeBaseVectorStoreDriver to query Griptape Cloud Knowledge Bases.
• GriptapeCloudEventListenerDriver.api_key defaults to the value in the GT_CLOUD_API_KEY environment variable.
• BaseObservabilityDriver as the base class for all Observability Drivers.
• DummyObservabilityDriver as a no-op Observability Driver.
• OpenTelemetryObservabilityDriver for sending observability data to an open telemetry collector or vendor.
• GriptapeCloudObservabilityDriver for sending observability data to Griptape Cloud.
• DatadogObservabilityDriver for sending observability data to a Datadog Agent.
• Observability context manager for enabling observability and configuring which Observability Driver to use.
• @observable decorator for selecting which functions/methods to provide observability for.
• GenericArtifact for storing any data.
• BaseTextArtifact for text-based Artifacts to subclass.
• HuggingFacePipelineImageGenerationDriver for generating images locally with HuggingFace pipelines.
• BaseImageGenerationPipelineDriver as the base class for drivers interfacing with HuggingFace image generation pipelines.
• StableDiffusion3ImageGenerationPipelineDriver for local text-to-image generation using a Stable Diffusion 3 pipeline.
• StableDiffusion3Img2ImgImageGenerationPipelineDriver for local image-to-image generation using a Stable Diffusion 3 pipeline.
• StableDiffusion3ControlNetImageGenerationPipelineDriver for local ControlNet image generation using a Stable Diffusion 3 pipeline.

Changed

• BREAKING: BaseVectorStoreDriver.upsert_text_artifacts optional arguments are now keyword-only arguments.
• BREAKING: BaseVectorStoreDriver.upsert_text_artifact optional arguments are now keyword-only arguments.
• BREAKING: BaseVectorStoreDriver.upsert_text optional arguments are now keyword-only arguments.
• BREAKING: BaseVectorStoreDriver.does_entry_exist optional arguments are now keyword-only arguments.
• BREAKING: BaseVectorStoreDriver.load_artifacts optional arguments are now keyword-only arguments.
• BREAKING: BaseVectorStoreDriver.upsert_vector optional arguments are now keyword-only arguments.
• BREAKING: BaseVectorStoreDriver.query optional arguments are now keyword-only arguments.
• BREAKING: EventListener.publish_event's flush argument is now a keyword-only argument.
• BREAKING: BaseEventListenerDriver.publish_event's flush argument is now a keyword-only argument.
• BREAKING: Renamed DummyException to DummyError for pep8 naming compliance.
• BREAKING: Migrate to sqlalchemy 2.0.
• BREAKING: Make sqlalchemy an optional dependency.
• BREAKING: Renamed drivers-sql-redshift to drivers-sql-amazon-redshift
• BREAKING: Renamed drivers-prompt-huggingface extra to drivers-prompt-huggingface-hub.
• BREAKING: Renamed drivers-vector-postgresql extra to drivers-vector-pgvector.
• BREAKING: Update marqo dependency to ^3.7.0.
• BREAKING: Removed drivers-sql-postgresql extra. Use drivers-sql extra and install necessary drivers (i.e. psycopg2) separately.
• Removed unnecessary sqlalchemy-redshift dependency in drivers-sql-amazon-redshift extra.
• Removed unnecessary transformers dependency in drivers-prompt-huggingface extra.
• Removed unnecessary huggingface-hub dependency in drivers-prompt-huggingface-pipeline extra.
• CsvRowArtifact now inherits from BaseTextArtifact.
• TextArtifact now inherits from BaseTextArtifact.

Fixed

• Parameter count for QdrantVectorStoreDriver.query now optional as per documentation.
• Path issues on Windows with LocalFileManagerDriver and AmazonS3FileManagerDriver.

0.28.2

Fixed

• Conversation Memory being incorrectly inserted into the PromptTask.prompt_stack when no system content is present.

0.28.1

Fixed

• Sending empty system content in PromptTask.
• Throttling issues with DuckDuckGoWebSearchDriver.

0.28.0

Added

• RagEngine is an abstraction for implementing modular RAG pipelines.
  • RagContext is a container object for passing around RAG context.
  • RAG stages:
    • QueryRagStage for parsing and expanding queries.
    • RetrievalRagStage for retrieving content.
    • ResponseRagStage for augmenting and generating outputs.
  • RAG modules:
    • Retrieval:
      • VectorStoreRetrievalRagModule for retrieving text chunks from vector stores.
      • TextLoaderRetrievalRagModule for retrieving data with text loaders in real time.
      • TextChunksRerankRagModule for re-ranking retrieved results.
    • Response:
      • MetadataBeforeResponseRagModule for appending metadata.
      • RulesetsBeforeResponseRagModule for appending rulesets.
      • PromptResponseRagModule for generating responses based on retrieved text chunks.
      • TextChunksResponseRagModule for responding with retrieved text chunks.
      • FootnotePromptResponseRagModule for responding with automatic footnotes from text chunk references.
• RagClient tool for exposing RagEngines to LLM agents.
• RagTask task for including RagEngines in any structure.
• Rerank drivers:
  • CohereRerankDriver for using the Cohere rerank API.
• utils.execute_futures_list() for executing a list of futures.
• LocalVectorStoreDriver.persist_file for persisting vectors and chunks in a text file.
• Entry.to_artifact() for easy vector store entry conversions into Griptape artifacts.
• BaseVectorStoreDriver.does_entry_exist() to check if an entry exists in the vector store.
• GoogleWebSearchDriver to web search with the Google Customsearch API.
• DuckDuckGoWebSearchDriver to web search with the DuckDuckGo search SDK.
• ProxyWebScraperDriver to web scrape using proxies.
• Parameter session on AmazonBedrockStructureConfig.
• Parameter meta on TextArtifact.
• VectorStoreClient improvements:
  • VectorStoreClient.query_params dict for custom query params.
  • VectorStoreClient.process_query_output_fn for custom query output processing logic.
• Parameter fail_fast to Structure.
• BooleanArtifact for handling boolean values.
• typos to dev dependencies to catch typos in code/docs.
• Message for storing messages in a PromptStack. Messages consist of a role, content, and usage.
• DeltaMessage for storing partial messages in a PromptStack. Multiple DeltaMessage can be combined to form a Message.
• TextMessageContent for storing textual content in a Message.
• ImageMessageContent for storing image content in a Message.
• Support for adding TextArtifacts, ImageArtifacts, and ListArtifacts to PromptStack.
• Support for image inputs to OpenAiChatPromptDriver, AzureOpenAiChatPromptDriver, AmazonBedrockPromptDriver, AnthropicPromptDriver, and GooglePromptDriver.
• Input/output token usage metrics to all Prompt Drivers.
• FinishPromptEvent.input_token_count and FinishPromptEvent.output_token_count.
• Support for storing Artifacts as inputs/outputs in Conversation Memory Runs.
• Agent.input for passing Artifacts as input.
• Support for PromptTasks to take TextArtifacts, ImageArtifacts, and ListArtifacts as input.
• Parameters sort_key and sort_key_value on AmazonDynamoDbConversationMemoryDriver for tables with sort keys.
• Reference for supporting artifact citations in loaders and RAG engine modules.

Changed

• BREAKING: Moved/renamed griptape.utils.PromptStack to griptape.common.PromptStack.
• BREAKING: Renamed PromptStack.inputs to PromptStack.messages.
• BREAKING: Moved PromptStack.USER_ROLE, PromptStack.ASSISTANT_ROLE, and PromptStack.SYSTEM_ROLE to Message.
• BREAKING: Updated return type of PromptDriver.try_run from TextArtifact to Message.
• BREAKING: Updated return type of PromptDriver.try_stream from Iterator[TextArtifact] to Iterator[DeltaMessage].
• BREAKING: Removed BasePromptEvent.token_count in favor of FinishPromptEvent.input_token_count and FinishPromptEvent.output_token_count.
• BREAKING: Removed StartPromptEvent.prompt. Use StartPromptEvent.prompt_stack instead.
• BREAKING: Removed Agent.input_template in favor of Agent.input.
• BREAKING: BasePromptDriver.run now returns a Message instead of a TextArtifact. For compatibility, Message.value contains the Message's Artifact value
• BREAKING: BaseVectorStoreDriver.upsert_text_artifact() and BaseVectorStoreDriver.upsert_text() use artifact/string values to generate vector_id if it wasn't implicitly passed. This change ensures that we don't generate embeddings for the same content every time.
• BREAKING: Removed VectorQueryEngine in favor of RagEngine.
• BREAKING: Removed TextQueryTask in favor of RagTask.
• BREAKING: TextArtifactStorage now requires vector_store_driver and rag_engine in place of vector_query_engine.
• BREAKING: Moved load_artifacts() from BaseQueryEngine to BaseVectorStoreDriver.
• BREAKING: Merged BaseVectorStoreDriver.QueryResult into BaseVectorStoreDriver.Entry.
• BREAKING: Replaced query_engine with vector_store_driver in VectorStoreClient.
• BREAKING: removed parameters google_api_lang, google_api_key, google_api_search_id, google_api_country on WebSearch in favor of web_search_driver.
• BREAKING: removed VectorStoreClient.top_n and VectorStoreClient.namespace in favor of VectorStoreClient.query_params.
• BREAKING: All futures_executor fields renamed to futures_executor_fn and now accept callables instead of futures; wrapped all future submit calls with the with block to address future executor shutdown issues.
• GriptapeCloudKnowledgeBaseClient migrated to /search api.
• Default Prompt Driver model in GoogleStructureConfig to gemini-1.5-pro.

Fixed

• CoherePromptDriver to properly handle empty history.
• StructureVisualizer.to_url() by wrapping task IDs in single quotes.