- Fix SQL injection vulnerabilities in chat_history.py, langchain_store.py, rag_pipeline.py, baradb_langchain.js by switching to parameterized queries - Replace dangerous bare except: with except CatchableError:/ValueError: in llm.nim, embed.nim, cypher.nim, mcp/server.nim, executor.nim - Fix session variable leak in MCP handleVectorSearch/handleSchemaInspect - Build: 0 errors, all tests pass
- clients/python/baradb/chat_history.py: BaraDBChatHistory class - Stores conversation threads in BaraDB with multi-tenant RLS - session_id + tenant_id + user_id isolation - Auto-creates table and index - Compatible with LangChain message format - examples/rag_pipeline.py: End-to-end RAG pipeline example - PDF/text ingestion -> chunking -> embedding -> BaraDB storage - Hybrid search with vector distance - LLM generation (OpenAI / Ollama) - Supports --file and --query modes - Configurable chunk size, overlap, top-k - PLAN.md: Updated all Session 10 tasks as complete
Cluster Auto-Rebalance: - ClusterMembership with onNodeJoin/onNodeLeave/onNodeFail - Automatic shard reassignment on node failure - Rebalance triggers on member changes Cypher-like Graph Queries: - MATCH/CREATE/MERGE/DELETE query parser - Full pattern parsing: nodes with labels/properties, edges with labels - WHERE, RETURN, ORDER BY, LIMIT clauses - executeCypher for basic graph matching Tutorial: - 8 comprehensive examples: KV store, B-Tree, vector search, graph, FTS, MVCC, BaraQL parsing, community detection - Runnable demo with all major engines 10 new tests (232 total, all passing)