HKUDS/DeepTutor · error · ValueError

Select at most one PageIndex OSS knowledge base per request.

Error message

Select at most one PageIndex OSS knowledge base per request. Selected: {', '.join(selected)}

What it means

validate_pageindex_oss_selection enforces that at most one PageIndex OSS (self-hosted) knowledge base may participate in a single request, because the OSS SDK binding is a singleton per process. Passing two or more distinct OSS KB names raises ValueError listing the offenders.

Source

Thrown at deeptutor/services/rag/pipelines/pageindex/__init__.py:41

        return None


def is_pageindex_kb(kb_name: str | None) -> bool:
    from deeptutor.services.rag.factory import PAGEINDEX_OSS_PROVIDER, PAGEINDEX_PROVIDER

    return _bound_pageindex_provider(kb_name) in {PAGEINDEX_PROVIDER, PAGEINDEX_OSS_PROVIDER}


def validate_pageindex_oss_selection(knowledge_bases: list[str]) -> None:
    from deeptutor.services.rag.factory import PAGEINDEX_OSS_PROVIDER

    selected = [
        name
        for name in dict.fromkeys(str(item or "").strip() for item in knowledge_bases)
        if name and _bound_pageindex_provider(name) == PAGEINDEX_OSS_PROVIDER
    ]
    if len(selected) > 1:
        raise ValueError(
            "Select at most one PageIndex OSS knowledge base per request. "
            f"Selected: {', '.join(selected)}"
        )


__all__ = ["PageIndexPipeline", "is_pageindex_kb", "validate_pageindex_oss_selection"]

View on GitHub (pinned to 3e82f13042)

Solutions

1. Remove all but one PageIndex OSS KB from the request; merge their documents into a single OSS KB if both corpora are needed.
2. If you need multiple corpora, use cloud PageIndex KBs alongside the single OSS one, or duplicate queries per OSS KB sequentially.

Example fix

# before
await pipeline.explore(query=q, knowledge_bases=["oss-main", "oss-archive"])

# after
await pipeline.explore(query=q, knowledge_bases=["oss-main", "cloud-kb"])

from deeptutor.services.rag.pipelines.pageindex import validate_pageindex_oss_selection
validate_pageindex_oss_selection(knowledge_bases)  # raises before any network work

def oss_kb_count(knowledge_bases: list[str]) -> int:
    from deeptutor.services.rag.pipelines.pageindex import _bound_pageindex_provider, PAGEINDEX_OSS_PROVIDER
    return sum(1 for n in dict.fromkeys(map(str, knowledge_bases)) if _bound_pageindex_provider(n) == PAGEINDEX_OSS_PROVIDER)

try:
    validate_pageindex_oss_selection(kbs)
except ValueError as e:
    if "at most one PageIndex OSS" in str(e):
        kbs = keep_single_oss_kb(kbs)  # dedupe to first OSS entry
    else:
        raise

When it happens

Trigger: Calling PageIndex pipeline explore()/handle() with a knowledge_bases list containing two or more names that bind to the PageIndex OSS provider (deduplicated, non-empty).

Common situations: Users selecting multiple OSS KBs in the UI multi-select, or API clients concatenating KB lists from different sources; cloud PageIndex KBs coexist fine, so the error fires only on multiple OSS entries.

Related errors

• PageIndex API key is not configured. Add it under Knowledge
• PageIndex OSS preflight failed: {details}
• No embedding model is configured. Set up the embedding profi
• Knowledge base '{resolved_name}' is not in an error state. U
• Knowledge base not initialized ({self.rag_provider}): {kb_na

AI-assisted analysis of HKUDS/DeepTutor@3e82f13042 (2026-08-27). Data as JSON: /api/errors/26c76ab9123e7842. Report an issue: GitHub.