v0.6.2 migrated to LlamaParse 2.x, backend in docker fixes

1. Migrated from using LlamaParse 1.x to using LlamaParse 2.x
2. Fixed default port for Neo4j to be 7687 if no .env
3. frontend package names fixed: now flexible-graphrag-react, flexible-graphrag-vue,
   already had flexible-graphrag-angular (had leftover names from cmis-graphrag-react, etc.)
4. For backend in docker: docker.env was fixed to have new style db specific json config vars,
   docker ontology paths fixed, access of postgres for inc update mgt fixed
5. In env-sample.txt, section with db config details in individual vars was removed,
   with details moved to json examples in regular db sections, added lm studio examples for
   openai_like llm and embedding.
6. for arcadedb embedded, readme.md and docs/getting-started/python-backend to have the correct
   package name arcadedb-embedded

v0.6.1: packaging fix (missing adapter files), pypi readme/description update, ontology docs fixes

v0.6.1: packaging fix (fix missing adapter files), pypi README/description improvements, ontology docs path corrections

1. flexible-graphrag/pyproject.toml — bump to 0.6.1; add adapters* to
   packages.find include (fixes ModuleNotFoundError: No module named
   'adapters' after PyPI install); rewrite description to be keyword-rich

2. flexible-graphrag/README.md — canonical intro paragraph; quick-install
   code block; LlamaIndex-only caveat with Framework Config doc link;
   corrected ontology schema paths (../schemas/); pointer to full README
   for Docker/frontend/extras

3. flexible-graphrag/start.py — fix docstring: uv run python -m start
   replaced with python -m start (former only works from source checkout)

4. flexible-graphrag-mcp/pyproject.toml — bump to 0.6.1; rewrite
   description (replace "HTTP bridge" with "connects to", add DB/source counts)

5. flexible-graphrag-mcp/README.md — top paragraph replaced with
   canonical project intro; remove "HTTP bridge" and GET /api/status
   references

6. docs/DATABASES/RDF/ontology_examples_and_config.md — clarify
   ontologies apply to both property graphs and RDF stores (not RDF-only);
   remove misleading RDF_GRAPH_DB block; correct all schema paths to
   ../schemas/ (matching env-sample.txt); add install-type note

7. CHANGELOG.md — 0.6.1 entry

v0.6.0 Full LangChain support in addition to LlamaIndex, more DBs (15 PG, 4 RDF)

v0.6.0 LlamaIndex or LangChain, 15 graph (8 + ArangoDB,AGE,Cosmos,SurrealDB,Spanner,HugeGraph,TigerGraph). 4 RDF (3 + Neptune RDF), 10 vector, 3 search

1. 15 total property graph databases — 8 existing LlamaIndex stores with LangChain versions
   added (Neo4j, ArcadeDB, FalkorDB, Memgraph, NebulaGraph, Neptune, Neptune Analytics,
   LadybugDB); 6 new LC-only stores (ArangoDB, Apache AGE, Azure Cosmos DB Gremlin, Apache
   HugeGraph, SurrealDB, TigerGraph); 1 new LI-only store (Google Cloud Spanner)

2. 10 LangChain vector backends (Qdrant, Elasticsearch, Milvus, Weaviate, LanceDB, Chroma,
   Pinecone, pgvector, OpenSearch, Neo4j vector); 3 LangChain search backends (Elasticsearch,
   OpenSearch, BM25); 4 RDF/triple-store backends (Fuseki, GraphDB, Oxigraph + new Amazon
   Neptune RDF with IAM SigV4 auth)

3. flexible-graphrag now runs fully on LlamaIndex, fully on LangChain, or any mix — both
   frameworks are first-class peers. Each pipeline stage is independently configurable:
   CHUNKER_BACKEND, KG_EXTRACTOR_BACKEND, GRAPH_BACKEND, VECTOR_BACKEND, SEARCH_BACKEND,
   RETRIEVAL_FUSION, LLM_PROVIDER, EMBEDDING_KIND. Note: document readers / data sources
   remain LlamaIndex-based (first pipeline stage). LangChain-only graph stores auto-select
   GRAPH_BACKEND=langchain.

4. Retrievers for both LI and LC with fusion support for both frameworks (RETRIEVAL_FUSION=
   llamaindex uses QueryFusionRetriever; =langchain uses EnsembleRetriever when all stores are
   LC-backed). LangChain retrievers include: Synonym Exploder (expands query terms for vector
   search), pg_vector + neighborhood traversal for Neo4j (LANGCHAIN_PG_VECTOR_SEARCH,
   USE_PG_NEIGHBORHOOD), and text-to-query graph QA for all LC property graph stores (generates
   Cypher for Neo4j/ArcadeDB/Memgraph/FalkorDB/Ladybug/AGE, GQL for HugeGraph, SurrealQL for
   SurrealDB, AQL for ArangoDB, SPARQL for all RDF stores).

5. Matrix test support — run_matrix.py / run_all_profiles.py; 24+ integration test profiles
   covering all PG, vector, search, RDF, and chunker combinations

6. Docling OCR — DOCLING_OCR=true + DOCLING_OCR_ENGINE (auto / rapidocr / easyocr /
   tesseract_cli / tesserocr / ocrmac); optional extras for easyocr, tesserocr, ocrmac

7. Incremental update (add, delete, modify) end-to-end across property graph, RDF,
   vector, and search databases on both LlamaIndex and LangChain backends

8. scripts/cleanup.py — all 15 property graph stores have native-client cleanup; early skip
   when store stage disabled to improve speed, postgres document state / datasource config
   tabvle cleanup skipped ifuse incremental update false.

9. Observability — upgraded OpenLIT + OpenInference LangChain instrumentation; both OTLP
   producers (LlamaIndex via OpenLIT, LangChain via OpenInference) work simultaneously

10. Docs site (zensical 0.0.40) + major doc updates — ARCHITECTURE.md (15 PG stores, 13 LLM
    providers, framework backends section), per-store setup guides (Cosmos Gremlin, Neptune,
    Spanner), CONFIG-PROPERTY-GRAPH, DATABASE-CONFIGURATION, UI-TAB-SEARCH, MCP-TOOLS; all
    broken links fixed, README.md updated

11. Per-store config isolation — each database and LLM/embedding provider has its own typed
    config env var ({TYPE}_GRAPH_DB_CONFIG, {TYPE}_VECTOR_DB_CONFIG, {TYPE}_SEARCH_DB_CONFIG,
    {KIND}_EMBEDDING_MODEL, etc.); per-store config takes precedence over generic fallback; no
    shared config collisions across stores

12. Time logging now separates out KG extraction time from graph storage time.

v0.5.2 ladybug pkg, py 3.14: MCP stdio and ingest text fixes

v0.5.2 Release Notes

v0.5.2 use ladybug package, llama-index-ladybug 0.3.1, py 3.14 MCP stdio fix, py 3.14 MCP/REST ingest_text, ingest_from_sample fix

Ladybug package rename

• Updated dependency from real-ladybug to ladybug>=0.15.3
• Updated dependency to llama-index-graph-stores-ladybug>=0.3.1
• factories.py import updated to match

Python 3.14 async fixes

• MCP stdio mode: removed nest_asyncio (broke on Python 3.14, was never needed)
• ingest_text / test_with_sample REST endpoints: call ingest/ingest_from_text: replaced sync refresh_ref_docs()
  with async_add() in ingest/ingest_from_text.py, fixing crash on Python 3.14;
  ingest_documents (used by UI) was unaffected as it already used the async path
• nest-asyncio removed from flexible-graphrag-mcp dependencies

v0.5.1 added LadyBug, modularized hybrid_system.py

v0.5.1 flexible-graphrag release notes

v0.5.1: Added Ladybug support, modularized large hybrid_system.py, removed Kuzu support, /rdf/schemas moved to project root

1. Added Ladybug support in config.py, factories.py, env-sample.txt, docker include for ladybug-explorer.yaml, added Ladybug info in /docs/
2. Removed Kuzu support in config.py, factories.py, env-sample, removed kuzu-explorer.yamal and kuzu-api.yaml docker includes, and remomved Kuzu info in /docs/
3. Split off code in hybrid_system.py into ingest/, process/, stores/, query_engine, retriever_setup, schema_manager; hybrid_system.py wires submodules.
4. FalkorDB ingest: patch falkordb.helpers and falkordb.graph stringify (bool, numpy, map keys) via llamaindex/graph/falkordb_param_patch.py.
5. docs/PERFORMANCE.md: legacy banner. will be completely redone; left Kuzu, will test Ladybug when redon; note about future split of KG extraction time with given LLM vs graph DB insert time.
6. File moving: cleanup.py and check_elasticsearch.py under moved to scripts/; Neptune wrapper under llamaindex/graph, removed unused cmis_util.py
7. README.md: Added LLM provider list to features section, Added Ladybug (Kuzu fork), removed Kuzu
8. requirements.txt removed — uv / pyproject.toml is the only supported install method
9. Bump flexible-graphrag and flexible-graphrag-mcp to v0.5.1.
10. flexible-graphrag-mcp/main.py — nest_asyncio.apply() moved out of HTTP mode; only applied in stdio mode to avoid anyio conflict on Python 3.14.
11. rdf/ontology_manager.py — relative ontology paths now resolve against os.getcwd() only
12. flexible-graphrag/rdf/schemas TTL files schemas now live at repo-root /schemas

v0.5.0 RDF ontologies, RDF Stores, 4 new LLM providers

Added support for 3 RDF stores, RDF ontologies, LangChain retrievers/llm, 4 new LlamaIndex LLM providers

1. Flexible GraphRAG now supports RDF-based ontologies for both property graph databases and RDF triple store databases
2. Added support for 3 RDF stores: Ontotext GraphDB, Fuseki, and Oxigraph.
3. The RDF stores support the same full pipeline as property graphs: document ingestion with KG extraction, auto incremental data source change detection, and UI search (hybrid search, AI query, and AI chat)
4. Add OWL/RDFS ontology-guided KG extraction (ontology_manager.py) — works with all property graph and RDF stores; URI map round-trip; XSD-typed literals from DatatypeProperty ranges
5. Add LangChain RDF QA fusion retriever (langchain/graph/) — TextToGraphQueryRetriever, SynonymExpander, GraphEntityVectorRetriever, GraphNeighborhoodRetriever
6. SynonymExpander can be configured on a retriever (LPG, RDF, etc) basis to improve LLM results
7. With SynonymExpander and GraphEntityVectorRetriever results from property graphs (initially with Neo4j)
8. Added native LangChain LLM factory (langchain/llm/) — all 13 providers with same config mapped to
   use LangChain LLMs
9. Added 4 new LLM providers on the LlamaIndex side: OpenRouter, LiteLLM proxy, openai_like, vLLM; scripts/litellm_config.yaml sample config
10. Added bundled ontology schemas (company_classes.ttl, company_properties.ttl, common_ontology.ttl, foaf_ontology.ttl); multi-ontology file support via ONTOLOGY_PATHS / ONTOLOGY_DIR
11. Added / updated docs: RDF-STORE-USER-GUIDE.md (implementation status, store dashboards), LANGCHAIN-GRAPH-INTEGRATION.md, LLM-EMBEDDING-CONFIG.md; README features, prerequisites, project structure;
12. Bumped version to 0.5.0

v0.4.0 ArcadeDB embedded mode added, pypi built releases (0.4.0)

ArcadeDB embedded mode added, pypi built releases (0.4.0) for flexible-graphrag and flexible-graphrag-mcp, readme updated, search fixes

1. ArcadeDB embedded mode: factories.py, config.py, env-sample.txt — runs in-process via ARCADEDB_MODE=embedded; optional HTTP/Studio via ARCADEDB_EMBEDDED_SERVER=true; arcadedb-embedded>=26.2.1 added as commented-out dependency
2. llama-index-graph-stores-arcadedb bumped to >=0.4.1 in requirements.txt and pyproject.toml for embedded seport
3. hybrid_system.py: X->REL->Y bare relation links filtered in all modes (not just graph-only); filtered before top_k slice
4. hybrid_system.py: Elasticsearch/OpenSearch-only ingestion error fixed
5. pyproject.toml: version 0.4.0; PEP 639 license; all 14 py-modules; incremental_updates package; Dockerfile/env-sample/requirements/uv.toml in sdist; placeholder README.md
6. start.py + pyproject.toml: flexible-graphrag console script entry point added
7. flexible-graphrag-mcp/pyproject.toml: version 0.4.0; Apache 2.0 license; author; readme reference
8. .gitignore: arcadedb_data/ and log/ excluded
9. README.md: badges added, backend pip install from pypi quickstart added; MCP server quickstart setup section added, mcp tools moved after; project/file structure updated; UI Usage moved after frontend setup; Testing cleanup mentions cleanup.py