From 3441783d5bc0934cea3046b18cb68b02dab1e190 Mon Sep 17 00:00:00 2001
From: Debanjum <debanjum@gmail.com>
Date: Sat, 30 Aug 2025 00:22:26 -0700
Subject: [PATCH] Disable web search tool if no search engine configured by
 admin

Webpage read is gated behind having a web search engine configured for
now. It can later be decoupled from web search and depend on whether
any web scrapers is configured.
---
 src/khoj/routers/helpers.py  |  3 +++
 src/khoj/routers/research.py |  5 +++++
 src/khoj/utils/helpers.py    | 11 +++++++++++
 3 files changed, 19 insertions(+)

diff --git a/src/khoj/routers/helpers.py b/src/khoj/routers/helpers.py
index 6c3eac0f..3bfbdb4c 100644
--- a/src/khoj/routers/helpers.py
+++ b/src/khoj/routers/helpers.py
@@ -121,6 +121,7 @@ from khoj.utils.helpers import (
     is_none_or_empty,
     is_operator_enabled,
     is_valid_url,
+    is_web_search_enabled,
     log_telemetry,
     mode_descriptions_for_llm,
     timer,
@@ -367,6 +368,8 @@ async def aget_data_sources_and_output_format(
             continue
         if source == ConversationCommand.Operator and not is_operator_enabled():
             continue
+        if source in [ConversationCommand.Online, ConversationCommand.Webpage] and not is_web_search_enabled():
+            continue
         source_options[source.value] = description
         if len(agent_sources) == 0 or source.value in agent_sources:
             source_options_str += f'- "{source.value}": "{description}"\n'
diff --git a/src/khoj/routers/research.py b/src/khoj/routers/research.py
index 43abaaf0..d242c5b9 100644
--- a/src/khoj/routers/research.py
+++ b/src/khoj/routers/research.py
@@ -36,6 +36,7 @@ from khoj.utils.helpers import (
     dict_to_tuple,
     is_none_or_empty,
     is_operator_enabled,
+    is_web_search_enabled,
     timer,
     tools_for_research_llm,
     truncate_code_context,
@@ -96,6 +97,7 @@ async def apick_next_tool(
         ConversationCommand.ViewFile,
         ConversationCommand.ListFiles,
     ]
+    web_research_tools = [ConversationCommand.SearchWeb, ConversationCommand.ReadWebpage]
     input_tools_to_research_tools = {
         ConversationCommand.Notes.value: [tool.value for tool in document_research_tools],
         ConversationCommand.Webpage.value: [ConversationCommand.ReadWebpage.value],
@@ -115,6 +117,9 @@ async def apick_next_tool(
         # Skip showing document related tools if user has no documents
         if tool in document_research_tools and not user_has_entries:
             continue
+        # Skip showing web search tool if agent has no access to internet
+        if tool in web_research_tools and not is_web_search_enabled():
+            continue
         if tool == ConversationCommand.SemanticSearchFiles:
             description = tool_data.description.format(max_search_queries=max_document_searches)
         elif tool == ConversationCommand.ReadWebpage:
diff --git a/src/khoj/utils/helpers.py b/src/khoj/utils/helpers.py
index e8c063d6..a43521d2 100644
--- a/src/khoj/utils/helpers.py
+++ b/src/khoj/utils/helpers.py
@@ -794,6 +794,17 @@ def is_internet_connected():
         return False
 
 
+def is_web_search_enabled():
+    """
+    Check if web search tool is enabled.
+    Set API key via env var for a supported search engine to enable it.
+    """
+    return any(
+        not is_none_or_empty(os.getenv(api_key))
+        for api_key in ["GOOGLE_SEARCH_API_KEY", "SERPER_DEV_API_KEY", "JINA_API_KEY", "FIRECRAWL_API_KEY"]
+    )
+
+
 def is_internal_url(url: str) -> bool:
     """
     Check if a URL is likely to be internal/non-public.