v7.0.0

[7.0.0] - 2025-01-23

Added

• Streaming responses feature that enhances QnABot responses by providing real-time streaming from Large Language Models (LLMs) to the chat interface. This introduces a cloudformation parameter EnableStreaming to optionally create resources needed for streaming through a nested stack. See README.
• Enhanced Guardrail Integration that implements pre-processing and post-processing guardrails to provide improved content control and broader security for your chatbot application. See README.
• Implemented Converse API to simplify LLM workflows by providing a consistent interface for different LLM providers, role prompting and eliminating the need for input tagging for Bedrock Guardrails. This introduces customizable system prompts LLM_GENERATE_QUERY_SYSTEM_PROMPT and LLM_QA_SYSTEM_PROMPT in content designer to support role-based prompting. For more information, see system prompts in supported models and model features and using Converse API.
• Ability to use both RAG with Bedrock KnowledgeBase and Kendra as fallback options. A new setting FALLBACK_ORDER in the content designer allows users to specify the fallback order of these options.
• Ability to set a TTL on records added to the DynamoDB UsersTable. (PR #671) - contributed by (@richhaase)
• Mistral as a new LLM provider option and support for latest Anthropic Sonnet 3.5 V2, Haiku 3.5 V1, Amazon Nova Models, Ai21 Jambda Instruct, Cohere R plus and Meta Llama 3.1 models.

Changed

• Upgraded to Node 20.
• Upgraded AWS SDK dependencies.
• Migrated Settings to DynamoDB store rather than SSM, allowing for longer custom settings and prompts.
• Moved Amazon Q Business Plugin from samples repo to QnABot repo as an example lambda hook
• Reduced the size of Bedrock KnowledgeBase output by removing citations from plaintext response.
• Updated OpenSearch EBSOptions VolumeType to gp3 from gp2.
• Updated IAM permissions and cloudformation outputs.
• Added additional non-user input fields to OpenSearch metrics data redaction exclusion list.
• Migrated to Poetry for Python dependency management
• Updated innerHTML usages to innerText per security best practices

Fixed

• Fixed issues with Excel file import and character handling when reading questions import file
• Improvements for merging of chained items (PR #720) - contributed by (@amendlik)

Deprecated

• Settings stored in SSM Parameters. These will automatically be moved to DynamoDB when you upgrade.
• Sagemaker embeddings and LLM workflows.
• KendraCrawlerSNS Topic workflow Issue #742
• Bedrock LLM Models Cohere Command Text, Jurassic-2 Mid and Ultra per Amazon Bedrock Model Lifefycle

Security

• Patched Jinja2, nanoid, path-to-regexp vulnerability

v6.1.5

[6.1.5] - 2024-11-20

Security

• Patched langchain, cross-spawn & elliptic vulnerability

v6.1.4

[6.1.4] - 2024-10-31

Fixed

• PII usage leaks and improvements. See README.

Security

• Patched http-proxy-middleware vulnerability

v6.1.3

[6.1.3] - 2024-10-17

Security

• Patched async and micromatch vulnerability

v6.1.2

[6.1.2] - 2024-10-07

Fixed

• Cleared context state credential and updated the page history after logout

Changed

• Added Anthropic Claude 3.5 Sonnet as an additional option to the list LLM models provided through cloudformation parameters LLMBedrockModelId and BedrockKnowledgeBaseModel

Deprecated

• Sagemaker support has been deprecated and will be removed in the next release

v6.1.1

[6.1.1] - 2024-09-26

Fixed

• Added back .gitignore to fix custom deployment issues through github repo
• Improved performance of lambda invocation from frontend to save settings faster
• Fixed bug that limited response card buttons to only 5 buttons Issue #765
• Security patch for body-parser, micromatch, path-to-regexp, and webpack
• Added support for crawled links in Bedrock Knowledge Base to be shown as referenced links
• Fixed an issue where the context is expanded by default and can't be closed when Knowledge Base returns lists in the response
• Fixed limit on import file sizes Issue #766

v6.1.0

[6.1.0] - 2024-08-29

Added

• Integration with Guardrails for Amazon Bedrock and Amazon Bedrock Knowledge Base Integration (see documentation)
• Ability to customize prompt template for RAG using Amazon Bedrock Knowledge Base through setting KNOWLEDGE_BASE_PROMPT_TEMPLATE (see documentation).
• Ability to customize inference parameters for LLM specified in BedrockKnowledgeBaseModel inference parameters for BedrockKnowledgeBaseModel through setting KNOWLEDGE_BASE_MODEL_PARAMS (see documentation)
• Ability to customize search type (e.g. SEMANTIC orHYBRID) for how data sources in the knowledge base are queried through setting KNOWLEDGE_BASE_SEARCH_TYPE (see documentation)
• Ability to customize maximum number of retrieved results for RAG using Amazon Bedrock Knowledge Base through setting KNOWLEDGE_BASE_MAX_NUMBER_OF_RETRIEVED_RESULTS (see documentation).
• Ability to customize metadata and filters for RAG using Amazon Bedrock Knowledge through setting KNOWLEDGE_BASE_METADATA_FILTERS (see documentation)
• Added an option to specify the retention period for log groups through cloudformation parameter LogRetentionPeriod
• Anonymized operational metrics for some designer settings

Changed

• Improved fault tolerance of Testall, Export, Import functionalities and added ContentDesignerOutputBucket
• Added Amazon Titan Text Embeddings V2 as an additional option to the list of embedding models provided through cloudformation parameter EmbeddingsBedrockModelId
• Added Amazon Titan Text Premier as an additional option to the list LLM models provided through cloudformation parameters LLMBedrockModelId and BedrockKnowledgeBaseModel. Issue 746
• Changed Sagemaker LLM image to latest
• Changed CustomQnABotSettings parameter store to Advanced Tier to accommodate storing additional custom settings

Removed

• Removed Amazon Lex V1 resources
• Removed Canvas LMS integration

Fixed

• Fixed import settings in content designer for double byte characters
• Fixed an edge case where the Knowledge Base could return a context starting with # characters, causing font differences in the returned text due to Markdown formatting
• Fixed session attribute qnabot_gotanswer not being set to true after receiving hits from Knowledge Base

Security

• Security patch for axios, moto, read-excel-file, handlebars, boto3, click, elliptic & postcss

v6.0.3

[6.0.3] - 2024-08-06

Security

• Patched fast-xml-parser vulnerability

v6.0.2

[6.0.2] - 2024-07-22

Added

• Added a migration documentation for Migrating QnABot configurations and data from existing deployment to new deployment
• Added a documentation for Bedrock Knowledge Base

Fixed

• Improve logout functionality which signs out the user and invalidates the access and refresh tokens that Amazon Cognito issued to a user. Issue #747
• Fixed bug that restricted import of questions with answers that consisted of only double-byte characters. Issue #731
• Fixed bug with chained questions causing errors in the fulfillment lambda.

Updated

• Removed aws-sdk (JavaScript V2) from dependency list.
• Updated parameter description for elicit response bot settings in the content designer settings. Issue #745
• Removed LLM models meta.llama2-70b-chat-v1 and meta.llama2-13b-chat-v1 from the list of models in the Cloudformation parameter LLMBedrockModelId since these models will be unavailable on Amazon Bedrock starting from August 12, 2024.
• Updated the setting LLM_QA_NO_HITS_REGEX in the Content Designer to include a default pattern Sorry, I don't know in prompts specified through the setting LLM_QA_PROMPT_TEMPLATE and other patterns returned by LLMs in their responses.
• Constrainted the query made to Bedrock Knowledge Base to maximum of 1000 characters input query as per the input requirements.

v6.0.1

[6.0.1] - 2024-06-26

Fixed

• Fixed bug that was restricting stack names to be below 26 characters. Issue #741
• Fixed a looping issue when using slots and chaining #721 (PR #721) - contributed by (@amendlik)
• Github links with incorrect paths.

Updated

• Security patches for braces, urllib3, and ws.
• Improved latency of IAM policy propagation when switching the Bedrock embedding model.