Messaging queues & vector databases

Cores under Messaging/ (solution Messaging/Messaging.sln) and VectorDatabase/. They still implement IDataSource, but “entities” often map to topics, queues, subscriptions, or collections—not relational tables. RDBMS CRUD/migration helpers usually do not apply.

IDataSource on messaging

  • Entity list may reflect topic/queue names or consumer group metadata, depending on the core.
  • Reads are often pull/poll or subscription-based; “current record” semantics differ from scrollable cursors.
  • Writes are publishes or enqueue operations; idempotency and ordering are broker-specific.
  • Kafka registers as DatasourceCategory.QUEUE; other messaging cores in this repo use MessageQueue.

Vector databases

Cores use DatasourceCategory.VectorDB. Operations center on collections/indexes, embedding dimensions, and similarity search rather than SQL SELECT. Upstream embedding models produce vectors; this repository’s datasources focus on storage and retrieval APIs.

In VectorDatabase/, some add-in attributes are commented out (e.g. Pinecone, SharpVector alternate entry types); the pages below reflect cores with an active [AddinAttribute] in source at documentation time.

Messaging (documented)

Vector (documented)

  • Qdrant — VectorDatabase/TheTechIdea.Beep.QdrantDatasource/ (QdrantDatasourceGeneric)
  • Milvus — VectorDatabase/TheTechIdea.Beep.MilvusDatasource/
  • ChromaDB — VectorDatabase/TheTechIdea.Beep.ChromaDBDatasource/

Contrast: product REST in Connectors/

Messaging and vector cores live under Messaging/ and VectorDatabase/ with queue/topic/collection semantics—not the CRM, e-commerce, or chat product APIs shipped under Connectors/<Category>/ as WebAPIDataSource. For Slack, Salesforce, Shopify-style HTTP operations, see REST / SaaS connectors and flagship conn-* index.

BeepDM skills

  • BeepDM/.cursor/idatasource/SKILL.md

See also: REST / SaaS connectors (overview and category hubs) and the flagship conn-* index.

Cloud & analytics Kafka