vector store migration to Chroma

[Buffden]

• rewrite ingest.py: chunk, embed, upsert into Chroma with source/chunk_index metadata
• rewrite search.py: replace brute-force cosine loop with collection.query() ANN search
• add inspect_collection.py: print collection stats and a sample stored entry
• move load_document/load_documents into utils.py; clean up embed.py to pure embedding utilities
• replace pipeline.svg with pipeline.puml (PlantUML source for Phase 2 diagram)
• add docs/implementation-plan.md: 7-phase build plan covering RAG, ingestion, hybrid search, chat mode, and access control
• update README: reflect Phase 2 stack, structure, usage, and sample output
• update pyproject.toml: rename project to rag-document-engine, add chromadb dependency
• update .gitignore: exclude chroma_db/ and *.dist-info/