SuperBizAgent-java/openspec/changes/archive/2026-07-05-rag-vectorstore-mainpath-with-sdk-fallback/tasks.md

1. Spring AI Milvus Setup

• 1.1 Add the official Spring AI Milvus VectorStore starter dependency.
• 1.2 Configure Spring AI Milvus to reuse the existing collection field names, dimension, metric type, and connection settings.
• 1.3 Add retrieval mode configuration for auto, spring-ai, and sdk.

2. Retrieval Service Refactor

• 2.1 Refactor VectorSearchService to inject optional Spring AI VectorStore.
• 2.2 Implement VectorStore search and map Spring AI Document results to existing SearchResult.
• 2.3 Preserve the existing SDK search path as a dedicated fallback method.
• 2.4 Route retrieval by configured mode and fallback rules.

3. Tests

• 3.1 Add tests for SDK mode bypassing VectorStore.
• 3.2 Add tests for auto mode using VectorStore when available.
• 3.3 Add tests for auto mode falling back to SDK when VectorStore fails or is unavailable.
• 3.4 Add tests for category metadata filter behavior.

4. Validation And Archive

• 4.1 Run targeted retrieval tests.
• 4.2 Run the offline RAG retrieval baseline evaluator.
• 4.3 Run OpenSpec validation.
• 4.4 Review git diff to confirm lookup_knowledge API and evidence output remain compatible.