stanford-oval/storm · error · RuntimeError

You must supply azure_ai_search_url or set environment varia

Error message

You must supply azure_ai_search_url or set environment variable AZURE_AI_SEARCH_URL

What it means

Raised by AzureAISearchRM.__init__ when the Azure AI Search service endpoint URL is neither passed as azure_ai_search_url nor found in the AZURE_AI_SEARCH_URL environment variable. The SDK needs the full service endpoint (e.g., https://<service>.search.windows.net) to build a SearchClient, so construction aborts.

Source

Thrown at knowledge_storm/rm.py:1157

            from azure.search.documents import SearchClient
        except ImportError as err:
            raise ImportError(
                "AzureAISearch requires `pip install azure-search-documents`."
            ) from err

        if not azure_ai_search_api_key and not os.environ.get(
            "AZURE_AI_SEARCH_API_KEY"
        ):
            raise RuntimeError(
                "You must supply azure_ai_search_api_key or set environment variable AZURE_AI_SEARCH_API_KEY"
            )
        elif azure_ai_search_api_key:
            self.azure_ai_search_api_key = azure_ai_search_api_key
        else:
            self.azure_ai_search_api_key = os.environ["AZURE_AI_SEARCH_API_KEY"]

        if not azure_ai_search_url and not os.environ.get("AZURE_AI_SEARCH_URL"):
            raise RuntimeError(
                "You must supply azure_ai_search_url or set environment variable AZURE_AI_SEARCH_URL"
            )
        elif azure_ai_search_url:
            self.azure_ai_search_url = azure_ai_search_url
        else:
            self.azure_ai_search_url = os.environ["AZURE_AI_SEARCH_URL"]

        if not azure_ai_search_index_name and not os.environ.get(
            "AZURE_AI_SEARCH_INDEX_NAME"
        ):
            raise RuntimeError(
                "You must supply azure_ai_search_index_name or set environment variable AZURE_AI_SEARCH_INDEX_NAME"
            )
        elif azure_ai_search_index_name:
            self.azure_ai_search_index_name = azure_ai_search_index_name
        else:
            self.azure_ai_search_index_name = os.environ["AZURE_AI_SEARCH_INDEX_NAME"]

View on GitHub (pinned to fb951af774)

Solutions

1. export AZURE_AI_SEARCH_URL="https://<your-service-name>.search.windows.net"
2. Or pass it explicitly: AzureAISearchRM(azure_ai_search_api_key=..., azure_ai_search_url="https://<svc>.search.windows.net", azure_ai_search_index_name=...)
3. Copy the exact 'Url' value from the Azure portal's Search service Overview page

Example fix

# before
rm = AzureAISearchRM(azure_ai_search_api_key=KEY, azure_ai_search_index_name=IDX)

# after
rm = AzureAISearchRM(azure_ai_search_api_key=KEY, azure_ai_search_url="https://mysvc.search.windows.net", azure_ai_search_index_name=IDX)

import os

def has_azure_search_url(explicit=None):
    return bool(explicit or os.environ.get("AZURE_AI_SEARCH_URL"))

assert has_azure_search_url(), "Missing AZURE_AI_SEARCH_URL"

try:
    rm = AzureAISearchRM(azure_ai_search_api_key=K, azure_ai_search_index_name=I)
except RuntimeError as e:
    if "AZURE_AI_SEARCH_URL" in str(e):
        raise SystemExit("Set AZURE_AI_SEARCH_URL to https://<service>.search.windows.net")
    raise

When it happens

Trigger: Constructing AzureAISearchRM with no azure_ai_search_url argument and no AZURE_AI_SEARCH_URL env var; reached only after the API key check has passed.

Common situations: Developer supplied the key but not the endpoint; copied the index name or resource name instead of the full endpoint URL; env var set in local shell but not in the deployed/containerized process.

Understand the failure class

Background: "environment variable is not set" and "Missing keys in environment" errors: what missing required env var messages mean and how to fix them — this error's family across 28 libraries.

Related errors

• You must supply azure_ai_search_api_key or set environment v
• You must supply azure_ai_search_index_name or set environmen
• You must supply google_cse_id or set the GOOGLE_CSE_ID envir
• Please provide an embedding model.
• Please provide a url for the Qdrant server.

AI-assisted analysis of stanford-oval/storm@fb951af774 (2026-08-28). Data as JSON: /api/errors/9f4ed93aa1c1c2d6. Report an issue: GitHub.