GraphForge v0.3.10 was an integration release.

After the TCK and performance push, this release added affordances that make GraphForge easier to embed in analytical, data pipeline, and agentic workflows.

What Shipped

• Schema introspection with labels(), relationship_types(), node_count(), and relationship_count()
• JSON export and import with roundtrip testing
• merge_node() safe upsert with create/match semantics
• add_graph_documents() for LangChain-compatible duck-typed ingestion
• Parse/plan LRU cache with cache_info() and clear_cache()
• Fixes for aliased RETURN DISTINCT, WITH-boundary variable reuse, and EXISTS edge cases
• Analytics integration documentation

Why We're Excited

This is the sort of release that makes a library easier to compose. Introspection, import/export, safe upsert, and document ingestion are all practical bridges between graph storage and the rest of a modern data system.

Source

• GitHub release: GraphForge v0.3.10