langchain-ai/langchain · error · ValueError

Vectorstore {destination} does not have required method {met

Error message

Vectorstore {destination} does not have required method {method}

What it means

Raised in `index()` during preflight checks when the destination `VectorStore` lacks a method required for indexing. Cleanup and upserts rely on `delete` and `add_documents`; if the object subclasses `VectorStore` but does not expose one of those methods (usually because a poorly-written subclass overrode or removed it, or a lazy proxy hides attributes), indexing is aborted with a `ValueError` naming the store and the missing method.

Source

Thrown at libs/core/langchain_core/indexing/api.py:436

    if (cleanup in {"incremental", "scoped_full"}) and source_id_key is None:
        msg = (
            "Source id key is required when cleanup mode is incremental or scoped_full."
        )
        raise ValueError(msg)

    destination = vector_store  # Renaming internally for clarity

    # If it's a vectorstore, let's check if it has the required methods.
    if isinstance(destination, VectorStore):
        # Check that the Vectorstore has required methods implemented
        methods = ["delete", "add_documents"]

        for method in methods:
            if not hasattr(destination, method):
                msg = (
                    f"Vectorstore {destination} does not have required method {method}"
                )
                raise ValueError(msg)

        if type(destination).delete == VectorStore.delete:
            # Checking if the VectorStore has overridden the default delete method
            # implementation which just raises a NotImplementedError
            msg = "Vectorstore has not implemented the delete method"
            raise ValueError(msg)
    elif isinstance(destination, DocumentIndex):
        pass
    else:
        msg = (  # type: ignore[unreachable]
            f"Vectorstore should be either a VectorStore or a DocumentIndex. "
            f"Got {type(destination)}."
        )
        raise TypeError(msg)

    if isinstance(docs_source, BaseLoader):
        try:
            doc_iterator = docs_source.lazy_load()

View on GitHub (pinned to e32fa9a52e)

Solutions

Implement the missing method on your store subclass (`add_documents` and `delete` are both mandatory for indexing).
Use `unittest.mock.create_autospec(VectorStore)` in tests so hasattr checks behave.
For search-only stores, do not use the indexing API — insert documents with the store's native path.

Example fix

# before
class ReadOnlyStore(VectorStore):
    # only similarity_search defined; no add_documents/delete
    ...

# after
class ReadOnlyStore(VectorStore):
    def add_documents(self, documents, **kwargs):
        self._client.upsert(...)
        return [d.id for d in documents]

    def delete(self, ids=None, **kwargs):
        self._client.delete(ids)
        return True

Defensive patterns

Strategy: type-guard

Validate before calling

required = ("add_documents", "delete")
missing = [m for m in required if not hasattr(store, m)]
if missing:
    raise TypeError(f"store lacks {missing}; cannot use indexing API")
index(store, docs, rm, cleanup="full")

Type guard

def supports_indexing(store) -> bool:
    return all(hasattr(store, m) for m in ("add_documents", "delete"))

Prevention

Implement the full required interface on custom VectorStore subclasses.
Use create_autospec(VectorStore) for test doubles so hasattr preflights behave.

When it happens

Trigger: Passing a custom VectorStore subclass that implements similarity search but not `add_documents` or `delete`; using `__getattr__`-based lazy proxies whose `hasattr` behavior is broken; partially-mocked stores in tests.

Common situations: Read-only vector store wrappers (search-only proxy over an external service); homegrown stores implementing only the retrieval interface; Mock objects without speccing.

Related errors

AI-assisted analysis of langchain-ai/langchain@e32fa9a52e (2026-08-14). Data as JSON: /api/errors/801774e993bf87c7. Report an issue: GitHub.