stanford-oval/storm · error · RuntimeError

You must supply google_cse_id or set the GOOGLE_CSE_ID envir

Error message

You must supply google_cse_id or set the GOOGLE_CSE_ID environment variable

What it means

Raised by GoogleSearchRM.__init__ (knowledge_storm/rm.py) when neither the google_cse_id constructor argument nor the GOOGLE_CSE_ID environment variable is set. The Google Custom Search Engine integration needs both an API key and a CSE (search engine) ID to issue queries, so the constructor fails fast before any search runs.

Source

Thrown at knowledge_storm/rm.py:1019

            k: Number of top results to retrieve.
            is_valid_source: Optional function to filter valid sources.
            min_char_count: Minimum character count for the article to be considered valid.
            snippet_chunk_size: Maximum character count for each snippet.
            webpage_helper_max_threads: Maximum number of threads to use for webpage helper.
        """
        super().__init__(k=k)
        try:
            from googleapiclient.discovery import build
        except ImportError as err:
            raise ImportError(
                "GoogleSearch requires `pip install google-api-python-client`."
            ) from err
        if not google_search_api_key and not os.environ.get("GOOGLE_SEARCH_API_KEY"):
            raise RuntimeError(
                "You must supply google_search_api_key or set the GOOGLE_SEARCH_API_KEY environment variable"
            )
        if not google_cse_id and not os.environ.get("GOOGLE_CSE_ID"):
            raise RuntimeError(
                "You must supply google_cse_id or set the GOOGLE_CSE_ID environment variable"
            )

        self.google_search_api_key = (
            google_search_api_key or os.environ["GOOGLE_SEARCH_API_KEY"]
        )
        self.google_cse_id = google_cse_id or os.environ["GOOGLE_CSE_ID"]

        if is_valid_source:
            self.is_valid_source = is_valid_source
        else:
            self.is_valid_source = lambda x: True

        self.service = build(
            "customsearch", "v1", developerKey=self.google_search_api_key
        )
        self.webpage_helper = WebPageHelper(
            min_char_count=min_char_count,

View on GitHub (pinned to fb951af774)

Solutions

1. Set the environment variable: export GOOGLE_CSE_ID="<your-cse-id>" (the 'Search engine ID' from programmablesearchengine.google.com)
2. Or pass it explicitly: GoogleSearchSearchRM(graph=graph, google_search_api_key=..., google_cse_id=...)
3. Ensure the variable is actually loaded (e.g., load_dotenv() before construction) if it lives in a .env file

Example fix

# before
rm = GoogleSearch(graph=graph, google_search_api_key=KEY)

# after
rm = GoogleSearch(graph=graph, google_search_api_key=KEY, google_cse_id=CSE_ID)
# or: export GOOGLE_CSE_ID=...

import os

def has_google_cse_config(explicit_cse_id=None):
    return bool(explicit_cse_id or os.environ.get("GOOGLE_CSE_ID"))

if not has_google_cse_config():
    raise SystemExit("Set GOOGLE_CSE_ID before starting the pipeline.")

try:
    rm = GoogleSearch(graph=graph, google_search_api_key=KEY)
except RuntimeError as e:
    if "GOOGLE_CSE_ID" in str(e):
        # load config from a secrets manager / .env and retry once
        load_dotenv()
        rm = GoogleSearch(graph=graph, google_search_api_key=KEY)
    else:
        raise

When it happens

Trigger: Instantiating GoogleSearchRM (or GoogleSearch via rm) without passing google_cse_id and without GOOGLE_CSE_ID exported in the environment. Note the preceding check also requires google_search_api_key / GOOGLE_SEARCH_API_KEY, so that must already be satisfied to reach this error.

Common situations: Developer set GOOGLE_SEARCH_API_KEY but forgot the separate CSE ID; env vars defined in .env but not loaded into the process; CI/容器 environments where the variable isn't exported; confusing the API key with the CSE ID (cx parameter) from the Programmable Search control panel.

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_url or set environment varia
• You must supply azure_ai_search_index_name or set environmen
• 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/c6b764ae80be1299. Report an issue: GitHub.