Changelog

v2.2 (2026-04-07)

Major Features

  • Agent Orchestrator (#739, #743, #744, #745, #746, #747, #748, #750): Multi-pattern agent orchestrator with LLM-based meta-routing to select the appropriate execution pattern per request:
    • Plan-then-Execute: LLM generates a plan of steps, executes each sequentially, and synthesises results
    • Supervisor: Decomposes a question into sub-agent goals, fans out to parallel sub-agents, aggregates findings into a synthesis
    • ReAct: Existing iterative reasoning pattern (unchanged)
    • Full explainability provenance for all patterns with new RDF types (Decomposition, Finding, Plan, StepResult, Synthesis) and predicates (tg:subagentGoal, tg:planStep)
    • Analysis split into Analysis+ToolUse and Observation for finer-grained DAG provenance; message_id wired on all streaming answer chunks
    • CLI support for pattern selection: tg-invoke-agent -p supervisor|plan-then-execute|react
  • RabbitMQ Pub/Sub Backend (#751, #752, #765): Pub/sub abstraction decoupled from Pulsar with RabbitMQ as an alternative backend, demonstrating independence from any single messaging fabric. RabbitMQ was selected for its significantly lower resource requirements compared to Pulsar. Support for additional fabrics such as Kafka is planned for a subsequent release.
    • Selectable via PUBSUB_BACKEND=rabbitmq environment variable
    • Topic exchange architecture with shared and exclusive consumer queues
    • Translator rename: to_pulsar/from_pulsarencode/decode across 55+ files
    • Queue naming format changed to CLASS:TOPICSPACE:TOPIC
    • Subscriber resilience: automatic consumer recreation after connection failure
    • Thread-safe consumer model with dedicated thread pools for pika
  • SPARQL Query Service (#754, #755): Backend-agnostic SPARQL 1.1 query service:
    • Parses SPARQL queries using rdflib, decomposes into triple pattern lookups via existing pub/sub interface
    • Supports BGP, JOIN, OPTIONAL, UNION, FILTER, BIND, VALUES, GROUP BY, ORDER BY, LIMIT/OFFSET, DISTINCT, and aggregates
    • Batching and streaming support for large result sets
    • Gateway integration, Python SDK method (FlowInstance.sparql_query), and CLI command (tg-invoke-sparql-query)
  • Universal Document Decoder (#705): Multi-format document processing using the unstructured library:
    • Supports DOCX, XLSX, PPTX, HTML, Markdown, CSV, RTF, ODT, EPUB and more through a single service
    • Tables preserved as HTML markup; images stored in librarian
    • Configurable section grouping strategies (whole-document, heading, element-type, count, size)
    • All decoders now share the document-decoder ident for interchangeability

Improvements

  • Inline Explainability Triples (#763): Provenance triples now included directly in explain messages from GraphRAG, DocumentRAG, and Agent services, eliminating follow-up knowledge graph queries for explainability details
  • Config Push Notify Pattern (#760): Replaced stateful pub/sub config broadcast with lightweight notify signal containing only version number and affected config types
  • Persistent WebSocket Connections (#723): Single persistent connection with request multiplexing replaces per-request WebSocket connections, eliminating repeated TCP+WS handshakes. CLI tools converted to concurrent WebSocket requests
  • Auto-pull Ollama Models (#757): Ollama provider automatically pulls missing models on first use
  • MCP Gateway Auth (#721): GATEWAY_SECRET environment variable support for MCP server to API gateway authentication
  • Chunk Content ID in Explain Traces (#708): tg-show-explain-trace now displays chunk URIs with --show-provenance for easy source text retrieval via tg-get-document-content
  • Prompt Queue Monitoring (#737): New tg-monitor-prompts CLI tool for subscribing to prompt request/response queues with correlation and timing summaries

Bug Fixes

  • Dispatcher Race Condition (#715): Fixed duplicate dispatcher creation under concurrent coroutines causing dropped responses and permanent UI spinners
  • WebSocket Error Responses (#726): Fixed missing request IDs in websocket multiplexer error responses causing client hangs on failed requests
  • OpenAI Compatibility (#727): Use max_completion_tokens instead of deprecated max_tokens for newer OpenAI/Azure models; added AZURE_API_VERSION environment variable override
  • Missing Auth Header (#724): Fixed verify_system_status processor check not including authorization header when gateway auth is enabled
  • Gateway Text Load (#729): Accept raw UTF-8 text in text-load endpoint
  • Stray Log Messages (#706): Removed spurious warnings from librarian responses arriving on shared response queues
  • Consumer Poll Timeout: Reduced consumer poll timeout from 2000ms to 100ms for improved responsiveness

Breaking Changes

  • Pub/sub queue naming: Queue format changed from topic-based to CLASS:TOPICSPACE:TOPIC; translator methods renamed from to_pulsar/from_pulsar to encode/decode
  • Agent schema: Orchestration fields added (correlation, sub-agents, plan steps); legacy response fields (answer, thought, observation) removed
  • Config push schema: ConfigPush now contains a types list instead of the full config dict; state queue class replaced by flow class

Infrastructure / Technical

  • Testing (#745, #749, #750): 96+ orchestrator tests covering aggregation, provenance, routing, explainability parsing, DAG structure, and callback message IDs
  • CLA Workflow (#716, #722): Contributor License Agreement process via GitHub action
  • Pulsar Check Skipped (#753): tg-verify-system-status no longer requires Pulsar when using alternative pub/sub backends

v2.1 (2026-03-17)

Major Features

  • Explainability & Provenance (#655, #661, #677, #682, #688, #689, #693, #694, #697, #698): End-to-end explainability across the entire pipeline:
    • Extract-time provenance: Document processing now emits PROV-O triples tracing the lineage from documents through pages, chunks, and extracted edges using prov:wasDerivedFrom relationships
    • Query-time explainability: GraphRAG, DocumentRAG, and Agent queries record full reasoning traces (question, grounding, exploration, focus, synthesis stages) into a dedicated urn:graph:retrieval named graph
    • Named graphs: Knowledge is now stored across named graphs — default graph for facts, urn:graph:source for extraction provenance, urn:graph:retrieval for query-time explainability
    • Subgraph provenance: Extracted subgraphs are tracked with provenance linking edges back to their source chunks and documents
    • New CLI tools: tg-list-explain-traces, tg-show-explain-trace, tg-show-extraction-provenance
    • Explainability modes added to tg-invoke-graph-rag, tg-invoke-document-rag, and tg-invoke-agent with inline provenance event display
  • Value to Term Schema Redesign (#622): Breaking redesign of the core wire format from Value ({"v": ..., "e": true}) to typed Term format:
    • IRIs: {"t": "i", "i": "http://..."}
    • Literals: {"t": "l", "v": "text", "d": "datatype", "l": "lang"}
    • Quoted triples (RDF-star): {"t": "r", "r": {"s": ..., "p": ..., "o": ...}}
    • Blank nodes: {"t": "b", "d": "identifier"}
    • Updated all processing pipelines, Cassandra indexes, serialization, and tests
  • Tool Services (#655, #656, #658): Dynamically pluggable tool implementations for agent frameworks:
    • Base class for creating custom tool services
    • Tool service client for the ReAct agent to discover and invoke tools at runtime
    • Tools can be deployed independently and registered dynamically
  • Batch Embeddings (#668, #669, #670, #671, #672, #681): Embeddings service redesigned for batch processing:
    • embed() now accepts a list of texts instead of a single text
    • Updated all embeddings providers (FastEmbed, Ollama, etc.)
    • Embeddings API now returns similarity scores
    • New CLI tools: tg-invoke-embeddings, tg-invoke-graph-embeddings, tg-invoke-document-embeddings, tg-invoke-row-embeddings

Improvements

  • Incremental / Large Document Loading (#659, #660): Multipart upload support for large documents:
    • S3 multipart upload with streaming retrieval
    • Upload session tracking in Cassandra with 24-hour TTL
    • New REST endpoint GET /api/v1/document-stream for streaming document content
    • New CLI tool: tg-get-document-content
  • Entity-Centric Graph (#633): Redesigned graph schema for entity-centric storage and querying
  • Structured Data Enhancements (#645, #646): Multi-index table support for structured data, removing need for manual Cassandra table modifications:
    • Row embeddings APIs exposed through gateway
    • New row-embeddings-query tool type for semantic search on structured data indexes
  • Streaming Triples (#676): Streaming triple queries with configurable batch sizes for lower time-to-first-result and reduced memory overhead:
    • tg-show-graph updated with --limit, --batch-size, --graph filter, and --show-graph options
  • Graph Query CLI (#679): New tg-query-graph tool for selective pattern matching on the triple store (by subject, predicate, object, graph) with auto-detection of value types
  • RDF-star Support in Turtle Export (#676): tg-graph-to-turtle now handles quoted triples and named graph filtering
  • Enhanced GraphRAG Pipeline (#691, #697): 4-stage GraphRAG pipeline with query concurrency and DocumentRAG grounding
  • Prompts JSONL Format (#619): Support for JSONL format in prompt definitions
  • Entity Context Enhancement (#629): Entity term now output alongside its definition in entity contexts
  • Terminology Rename (#682): Clarified naming throughout — “provenance” callbacks/IDs renamed to “explain” for clarity

Bug Fixes

  • Cassandra Schema and Graph Filter Semantics (#680): Fixed Cassandra schema for named graph support and corrected graph filter semantics
  • Subscriber Queue Clogging (#642): Fixed unexpected messages causing subscriber queue clogging
  • Google AI Studio (#641, #639, #640): Fixed Google AI Studio integration, moved to VertexAI package to simplify dependencies
  • VertexAI SDK Migration (#632): Migrated from deprecated Google GenAI library to the google-genai SDK
  • LLM Metrics (#631): Fixed metric label issues across LLM providers
  • Azure LLM Model (#657): Fixed model parameter usage in Azure LLM integration
  • Ontology URI Issue (#637): Fixed ontology URI handling
  • Entity/Triple Batch Size Limits (#635): Added batch size limits to prevent oversized requests
  • Pipeline Metadata ID Overwrite (#686): Fixed metadata id field being overwritten at each processing stage
  • Null Embeddings Protection (#627): Added guard against null embeddings
  • Graph Embeddings Service Identifier (#648): Fixed mismatching ge-query / graph-embeddings-query service identifiers
  • Rate Limiting (#638): Use ClientError and status code to correctly detect 429 rate-limit errors
  • Mistral SDK (#687): Locked mistralai to <2.0.0 to avoid a breaking change
  • KG Extraction (#695): Removed schema:subjectOf edges from knowledge graph extraction

Breaking Changes

  • Value to Term wire format: All API clients must update to the new Term format (see Major Features above)
  • tg-invoke-objects-query renamed to tg-invoke-rows-query; gateway service key changed from objects to rows
  • tg-load-pdf and tg-load-text removed: Document loading is now handled through the library/processing pipeline
  • Metadata field: metadata.metadata (subgraph) replaced by metadata.root (simple value) in export/import serialization
  • Embeddings fields: vectors (plural) became vector (singular); document embeddings now reference chunk_id instead of inline chunk text
  • Graph store: Only Cassandra is currently implemented as a graph store backend. Neo4j and Memgraph support is not available in this release.
  • Vector store: Only Qdrant is currently implemented as a vector store backend. Milvus support is not available in this release.

Infrastructure / Technical

  • Tech Specs: Added technical specifications for agent explainability, tool services, graph contexts, extraction dataflow, and structured data multi-index
  • Testing (#647, #663, #666, #696): Updated and expanded test suite for new Term schema, explainability, provenance, and embeddings interfaces

v1.8 (2026-01-19)

Major Features

  • API Documentation (#612, #613, #614): Comprehensive API specifications and documentation:
    • REST API OpenAPI specification with full endpoint coverage
    • WebSocket AsyncAPI specification for real-time interactions
    • Python API documentation with auto-generation tooling
    • Removed legacy hand-written API documentation in favor of generated specs
  • Messaging Fabric Plugins (#592): Plugin architecture for messaging fabric enabling alternative messaging backends:
    • Technology-neutral schema expressions for transport abstraction
    • Backend abstraction layer for pub/sub operations
    • Enables future support for messaging systems beyond Pulsar
    • Schema strictness improvements uncovered and fixed incorrect schema usage

Improvements

  • Generic S3 Storage Support (#594): Librarian blob storage refactored for S3-compatible stores:
    • MinIO-specific options changed to generic S3 parameters
    • Added region and SSL configuration options
    • Integrated with Garage - the configuration portal delivers integrated Garage
  • Storage Management Cleanup (#595): Addressed legacy issues in storage management:
    • Removed legacy storage management code
    • Fixed deletion of last collection edge case
    • Storage processors now ignore data for deleted collections
  • URL Normalization (#617): Gateway URLs now work with or without trailing slashes

Bug Fixes

  • Configuration Fixes (#616, #609, #611, #610): Multiple config-related fixes:
    • Fixed flows/flow key issue in config service
    • Fixed config inconsistency issues
    • Fixed flow loading problems
    • Fixed load-doc command issues
  • Streaming Fixes (#607, #608, #602, #599): Resolved streaming-related issues:
    • Fixed non-streaming RAG problems
    • Fixed agent streaming tool failure
    • Fixed various streaming API issues
  • Schema Fixes (#598, #596): Schema message improvements:
    • Fixed doc embedding schema messages
    • Fixed optionality in objects-query schema
  • Collection Management (#597): Fixed collection existence test logic
  • Dependencies (#606): Added missing trustgraph-base dependency

Testing

  • Streaming Tests (#600, #601): Added comprehensive streaming tests and fixed async test warnings

v1.7 (2025-12-23)

Major Features

  • Multi-Tenant Support (#583): Basic multi-tenant infrastructure enabling isolated deployments:
    • Collection management migrated to config service from librarian
    • Fixed parameter name mismatches for queue customization
    • Collection storage now uses config service with push-based distribution
    • Fixed AsyncProcessor and Config Service parameter handling
    • Services can now use tenant-specific queues and configurations
  • Python API Refactor (#577): Comprehensive Python API client enhancement with feature parity and streaming support:
    • Streaming interfaces for all LLM services (agent, GraphRAG, DocumentRAG, text completion, prompts)
    • WebSocket transport for persistent connections and multiplexing
    • Async/await support across all interfaces (REST, WebSocket, bulk, metrics)
    • Bulk import/export for triples, graph embeddings, and document embeddings
    • 60x latency improvement for streaming operations (500ms vs 30s first token)
    • Type-safe interfaces with full backward compatibility
    • CLI utilities updated to use new streaming API
  • Improved Ontology Extraction (#576): Enhanced ontology-based knowledge extraction:
    • Entity normalizer for consistent entity naming
    • Simplified parser for improved extraction accuracy
    • Triple converter for better schema adherence
    • Enhanced prompt engineering for ontology extraction

Improvements

  • System Monitoring (#579): System startup tracker for deployment verification with CLI tool tg-verify-system-status
  • Logging Enhancements (#586, #588): Production-grade logging infrastructure:
    • Loki logging integration for centralized log aggregation
    • Service ID added to log entries instead of module name
    • Enhanced logging strategy with structured output
  • Metrics (#589): Added model information to metering metrics for better cost tracking and analysis
  • Gateway Configuration (#584): Gateway queue overrides for flexible deployment topologies

Infrastructure / Technical

  • Tech Specs: Added comprehensive technical specifications:
    • Multi-tenant support architecture
    • Python API refactor design
    • Ontology extraction phase 2
    • Enhanced logging strategy
  • Testing: Added comprehensive Python API client tests with streaming validation

v1.6 (2025-12-04)

Major Features

  • Streaming LLM Responses (#566, #567): Comprehensive streaming support for LLM text completion enabling real-time token-by-token delivery:
    • Infrastructure with streaming flag in schemas, Gateway API (REST/WebSocket), Python API, and CLI tools.
    • Full streaming implementation across all LLM providers including Azure, Azure OpenAI, Bedrock, Claude, Cohere, Google AI Studio, Llamafile, LM Studio, Mistral, Ollama, OpenAI, TGI, Vertex AI, and vLLM
    • Backward compatible with existing non-streaming clients
    • Support for WebSocket streaming
    • Reduces time-to-first-token and improved UX for long responses
  • Streaming RAG Responses (#568): Extended streaming support to GraphRAG and DocumentRAG services:
    • Token-by-token responses for knowledge graph and document retrieval queries
    • Consistent streaming UX across all TrustGraph services
    • Leverages existing PromptClient streaming infrastructure
    • Gateway support via WebSocket for real-time client applications
  • Streaming Agent Interactions (#570): Enhanced agent framework with streaming support:
    • Real-time streaming of ReAct agent thought/observation/answer chunks
    • Incremental response delivery for multi-step agent workflows
    • Streaming parser for agent responses with robust error handling

Improvements

  • Enhanced Integration Tests (#568, #570): Comprehensive test coverage for streaming functionality

Bug Fixes

  • AWS Bedrock Model Invocation (#572): Fixed compatibility issues with newer Bedrock model invocation API including proper streaming support
  • Minio Library Compatibility (#565): Fixed incompatible library change in Minio client for blob storage operations
  • Streaming Agent Interactions (#570): Fixed race conditions and message ordering issues in streaming agent responses

Infrastructure / Technical

  • CLI Improvements: Enhanced CLI tools with streaming output:
    • tg-dump-queues: New utility for developer queue diagnostics

Templates

  • Updates to Bedrock and Claude models to support latest models

v1.5 (2025-11-23)

New Features

  • OntoRAG: Ontology-Based Knowledge Extraction (#523): New processor kg-extract-ontology that uses ontology objects from config to guide triple extraction. Includes entity contexts and integrates with ontology extractor from workbench.
  • MCP Authentication (#557): Added MCP auth token header support for the simple authentication case.
  • Dynamic Embeddings Model Selection (#556): Embeddings model can now be selected dynamically rather than being fixed.

Bug Fixes

  • Collection deletion batch error (#559): Fixed batch error during collection deletion by reducing batch size.
  • Fix hard-coded vector size (#555): Vector store now lazy-creates collections with different collections for different dimension lengths.
  • Fix AgentStep schema error (#557): Agent step argument values are now converted to strings to fix schema errors.
  • Remove unnecessary OpenAI parameters (#561): Removed parameters from OpenAI invocation that were causing compatibility issues with OpenAI-compatible services.

Infrastructure / Technical

  • Python 3.13 Support (#553): Upgraded to Python 3.13, switched from cassandra-driver to scylla-driver (cassandra-driver doesn’t work with Python 3.13).
  • Vector Store Lifecycle Tech Spec (#555): Added technical specification for vector store lifecycle.
  • OntoRAG Tech Spec (#523, #558): Added technical specification for ontology-based knowledge extraction and query.

Testing

  • Ontology extraction tests (#560): Added tests for ontology extraction.
  • Dynamic embeddings tests (#556): Added tests for dynamic embeddings model selection.
  • MCP auth and agent step parsing tests (#557): Added tests for MCP authentication and agent step parsing.
  • OpenAI invocation tests (#561): Updated tests for OpenAI parameter changes.

v1.4 (2025-10-06)

New Features

  • Flow Configurable Parameters (#526, #530, #531, #532, #533, #541): Major enhancements to flow parameter system:
    • Flow configurable parameters with dynamic settings
    • LLM dynamic settings using llm-model and llm-rag-model parameters
    • Temperature parameter support for all LLMs
    • Flow creation uses parameter defaults in API and CLI
    • Advanced parameter mode with controlled-by relationships
    • New CLI tools: tg-show-parameter-types
    • Dynamic chunking parameters
  • Structured Data Diagnosis Service (#518, #519):
    • New structured data diagnosis service plumbed into API gateway
    • Added XML, JSON, CSV detection capabilities
    • Type detector with schema selection
  • Enhanced Collection Management (#520, #522, #542, #544):
    • Collection metadata management and deletion capabilities
    • Librarian services integrated with collection manager
    • Collection tracking across all processors
    • Explicit collection creation/deletion (removed implicit creation)
    • Fixed collection management synchronization issues
  • User/Collection Isolation (#509, #510):
    • Neo4j user/collection separation
    • Memgraph user/collection processing

Improvements

  • Cassandra Performance (#521):
    • Refactored Cassandra knowledge graph for single table
    • Multi-table implementation for performance enhancement
    • Added Cassandra collection table
  • GraphRAG Optimizations (#527): Implemented GraphRAG optimizations with updated tests
  • Vector Store Enhancements (#512): Vector stores now create collections on query
  • Build System (#515): Parallel container builds
  • Logging (#528, #543): Reduced excessive request/response logging and log spam

Bug Fixes

  • Collection Management (#544): Fixed collection management synchronization problems
  • Metrics (#539, #540): Fixed label names and label issues in metrics
  • WebSocket (#517): Fixed async websocket closure handling
  • CLI (#529): Fixed CLI typo
  • Tests (#534, #535): Fixed failing tests and improved LLM parameter test coverage
  • Object Writer (#544): Fixed object writer management issues
  • Milvus (#544): Updated Milvus to use ANN correctly

API Changes

  • Gateway (#514): Return empty embeddings list as empty list through gateway.
  • Parameters: Changed parameters to parameter-types for consistency

v1.3.0

New Features

  • Structured Data Enhancements (#492, #493, #496, #498, #500): Major improvements to structured data handling:
    • NLP query to GraphQL service for natural language database queries
    • Structured query tool integration with agent framework
    • Enhanced structured query support and object batching
    • Structured data loader CLI with auto mode functionality
    • Object import capabilities with batch processing
  • Collection Management (#503, #511):
    • Extended use of user + collection fields throughout the system
    • Stores automatically create collections on query
  • Tool Groups (#484): Added tool grouping functionality for better organization

Improvements

  • GraphQL Enhancements (#486, #489):
    • Added GraphQL table query support
    • Removed redundant GraphQL collection parameter
  • Cassandra Configuration Standardization (#483, #488, #490):
    • Made Cassandra options (user, password, host) consistent across all utilities
    • Consolidated Cassandra configuration for better consistency
    • Refactored Cassandra operations to use common helper functions
  • API Improvements (#513): Return empty embeddings list as empty list through gateway

Bug Fixes

  • Vector Store Fixes (#507): Fixed Milvus vector store integration issues
  • Document Processing (#506): Fixed document RAG processing issues
  • Monitoring (#502): Fixed Prometheus incorrect metric names
  • API Consistency (#481): Fixed trustgraph-base chunks/documents confusion in the API
  • System Integration (#494): Resolved various system integration issues
  • Import/Export (#476): Fixed graceful shutdown for import/export operations
  • Knowledge Loading (#472): Use collection field from request when loading knowledge core

v1.2.17

New Features

  • MCP Tool Arguments Support (#462): Added support for Model Context Protocol (MCP) tool arguments, including agent support and additional tests
  • Anthropic Support for VertexAI (#458): Added Anthropic model support for Google Vertex AI platform
  • Knowledge Load Utility CLI (#456): New command-line utility for loading knowledge into the system
  • Structured Data MVP (#452): Initial implementation of structured data handling with:
    • New schemas and architecture principles
    • Object extractor functionality
    • Cassandra object storage support
  • Knowledge Extraction via Agent Framework (#439):
    • Implemented KG extraction agent (kg-extract-agent)
    • Using ReAct framework (agent-manager-react)
    • Refactored ReAct manager to use traditional ReAct messages

Improvements

  • Agent Tool Coverage (#460): Increased ReAct tool coverage with multi-step tool invocation and reasoning tests
  • Schema Structure Refactor (#451): Major refactoring of schema structure for better organization
  • Logging Strategy (#444): Implemented comprehensive logging strategy, converting all print() statements to proper logging invocations
  • Build System Modernization (#440): Migrated from setup.py to pyproject.toml for modern package infrastructure

Bug Fixes

  • Agent Tool Resilience (#461):
    • Fixed incorrect tool initialization in agent service
    • Made Action parsing more resilient by handling quotation marks
  • Missing Anthropic Import (#459): Fixed missing import for Anthropic functionality
  • Token Chunker API (#454, #455): Fixed broken API invocation in token chunker
  • Librarian Collection Validation (#453): Added validation for librarian collections
  • Mistral OCR (#448, #450):
    • Fixed Mistral OCR to use current API
    • Corrected Mistral OCR identifier to standard pdf-decoder
  • Logging Startup Issues (#445, #446, #447): Resolved multiple logging startup problems

Infrastructure

  • Build Dependencies (#441, #442): Added missing build dependencies
  • Template Addition (#463): Added new template support
  • Python Dependencies: Updated Python dependencies to version 1.2

Testing

  • PDF Decoder Tests: Added comprehensive tests for PDF decoder functionality
  • MCP Arguments Tests: Added test coverage for MCP tool arguments
  • Multi-step Reasoning Tests: New tests for multi-step tool invocation scenarios

v1.1.10

New Features

  • MCP (Model Context Protocol) Support: Added MCP server and client support (#419, #425, #426, #427, #428)
  • React Integration: Added React call MCP functionality (#428)

Improvements

  • Documentation updates for API/CLI changes in v1.0 (#420, #421)
  • Enhanced README with messaging improvements and link fixes

v1.0.22

Major Features

  • Flow API & Management: Complete flow configuration and management system (#345, #346, #356, #357, #358)
  • Knowledge Management: Knowledge service, library management, and core CLI tools (#367, #368, #369, #372)
  • Enhanced Gateway: Reverse gateway functionality and improved API gateway (#416, #356)
  • Performance Improvements: Multi-threading support for consumers and LLMs (#408, #409)

New Integrations

  • vLLM Support: Added vLLM integration (untested) (#410)
  • HuggingFace TGI: Added HuggingFace Text Generation Inference support (#396)
  • Google AI Updates: Enhanced Google AI integration (#394)

Bug Fixes & Improvements

  • Fixed command line arguments handling (#417)
  • Fixed library translators (#415)
  • Fixed missing script issues (#418)
  • Improved token rate measurement utility (#412)
  • Enhanced translator classes (#414)
  • Miscellaneous fixes (#413)

v0.23

New Features

  • Knowledge Service: Full knowledge management system with CLI tools
  • Library Management: Document submission and library CLI functionality
  • Entity Contexts: Import/export capabilities for entity contexts
  • Configuration Persistence: Enhanced config management and reload mechanisms

Infrastructure

  • Container Updates: Upgraded to Python 3.12 (#386)
  • OCR Improvements: Ported OCR code to new API architecture
  • Apache 2 License: Updated licensing (#373)

Bug Fixes

  • Fixed LLM launch bugs (#377, #378)
  • Fixed queue initialization issues (#381, #382)
  • Fixed chunking not being enabled (#364)
  • Fixed OpenAI base URL handling

v0.22

Major Features

  • Configuration Service: Dynamic configuration management (#332, #334, #335)
  • Prompt Management: Dynamic prompt loading and CLI commands (#338)
  • Agent Management: Enhanced agent configuration system

Infrastructure Changes

  • Removed template directory (moved to separate repository) (#336)
  • Enhanced configuration initialization (#335)
  • Added config reload handler mechanism (#334)

v0.21

New Features

  • Librarian Service: Document processing and management (#304, #305, #307, #310, #311)
  • Mistral AI Support: Complete Mistral API integration including OCR (#313, #326)
  • LM Studio Integration: Added LM Studio LLM hosting support (#323, #328)
  • PDF OCR: Separate PDF OCR package with Tesseract support (#324)
  • Cassandra Integration: Added Cassandra auth with SSL support (#318)

Cloud & Infrastructure

  • Azure AKS: Azure Kubernetes Service integration (#317, #319)
  • AWS Bedrock: Inference profiles support (#314)
  • FastEmbed: Enabled FastEmbed component (#316)

API Improvements

  • Pulsar API Key: Added API key support (#308)
  • Async/Sync Fixes: Resolved async/sync loading issues (#315)
  • GraphRAG & DocRAG: Enhanced parameters and path hops (#312)

Bug Fixes

  • Fixed broken setup.py (#320)
  • Fixed async send typos (#322)
  • Fixed container build issues (#325)
  • Fixed missing OpenAI symbol and base URL specification (#330)
  • Fixed Bedrock integration issues (#331)