haystack · Issues· 155 open

DocumentSplitter emits redundant overlap-only trailing chunks for character-based split modes (token mode was fixed in #12661)

Agent generation_kwargs["tools"] replaces agent.tools instead of being treated as client-side specs

CSVToDocument row mode overwrites a CSV column named row_number

CSVDocumentCleaner deletes a row that says N/A, and both CSV components empty those cells

OpenAIResponsesChatGenerator discards annotations

CSVDocumentCleaner silently removes nonempty NA and NULL strings

JsonSchemaValidator raises IndexError for an empty message list

JsonSchemaValidator rejects valid empty schemas and ignores empty runtime overrides

RecursiveDocumentSplitter split_idx_start is off by the trailing separator with split_unit="word" and split_overlap > 0

Agent tool scheduler picks write-write winner by batch order, not call order, contradicting its own docstring and #11690

4 of 5 Document*Evaluators crash with ZeroDivisionError on empty input

OutputAdapter silently coerces string output to another type when output_type is a Union including str.

Bug: MarkdownHeaderSplitter shares parent_headers list across secondary chunks when keep_headers=False

InMemoryDocumentStore unique metadata count fails for list and dict values

CacheChecker.run issues one filter_documents call per item (N+1 query pattern)