feat: update adapters for MERGE edge support + wildcard patterns + vector search

[polaz]

CoordiNode v0.3.1 shipped three fixes that unblock SDK improvements:

1. Wildcard relationship patterns — MATCH (n)-[r]->(m) now returns results.

   • LlamaIndex get_triplets() can drop the NotImplementedError for missing relation_names
   • LlamaIndex get_rel_map() can use wildcard instead of schema-lookup typed pattern
   • LangChain refresh_schema() can use wildcard instead of typed pattern

2. MERGE for edge patterns — MERGE (src)-[r:TYPE]->(dst) works.

   • LlamaIndex upsert_relations() and LangChain _create_edge() can switch from CREATE to MERGE for idempotent edge upsert
   • LangChain _link_document_to_entities() MENTIONS edges can use MERGE

3. Vector search via VectorService RPC — returns real results.

   • E2E integration tests needed for vector_search() / vector_query()

Also clean up stale workaround comments and docstrings that reference now-fixed limitations.

[coderabbitai]

⚠️ Possible Duplicate Issue(s)

• fix: duplicate edges on re-ingestion — replace CREATE with MERGE in upsert_relations() and add_graph_documents() #21

🔗 Related PRs

#15 - fix: CoordiNode Cypher compatibility — add_graph_documents, type/label, MATCH+CREATE [merged]
#18 - build(deps): bump minimum coordinode requirement to >=0.4.0 [merged]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!