I performed a quick investigation of the Langchain documentation and came up with the following.

Issues found in the Integrations section

Retrievers:

• MongoDB only has two retrievers listed (fulltext & hybrid).. It is missing: Parent-Document, Self-Querying, GraphRAG
• Bring-your-own documents: At the top of this page, a small group are singled out “The below retrievers allow you to index and search a custom corpus of documents.” We are not in that group.

Tools: Database section is missing MongoDBDatabaseToolkit.

VectorStores: Ours is shown to fail the following: [standard tests, multitenancy, IDs in add documents]. We have all of these.

API documentation

The API documentation is performed by us. The link from their site to ours works.

Definition of done

How are these created? By someone at LangChain manually or via CI? These are small enough that we can do them ourselves if the code is available and submit a pull-request.