feat(community): add Serper web search provider (#2630)

* feat(community): add Serper web search provider

Add a new community search provider backed by the Serper Google Search
API (https://serper.dev). Serper returns real-time Google results via a
simple JSON API and requires only an API key — no extra Python package.

Changes:
- backend/packages/harness/deerflow/community/serper/__init__.py
- backend/packages/harness/deerflow/community/serper/tools.py
  Implements web_search_tool using httpx (already a project dependency).
  API key is read from config.yaml `api_key` field or SERPER_API_KEY env var.
  Follows the same interface / output shape as the existing ddg_search provider.
  Exposes max_results parameter (default 5) with config override logic.
- backend/tests/test_serper_tools.py
  Unit tests covering API key resolution, config overrides, HTTP errors,
  empty results, and parameter passing.
- config.example.yaml: add commented-out Serper example alongside other providers
- .env.example: add SERPER_API_KEY placeholder

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

* Fix the lint error

* Fix the lint error

---------

Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Co-authored-by: Willem Jiang <willem.jiang@gmail.com>
This commit is contained in:
wanxsb 2026-05-02 16:22:35 +08:00 committed by GitHub
parent e543bbf5d6
commit 44ab21fc44
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
5 changed files with 419 additions and 0 deletions

View File

@ -1,3 +1,6 @@
# Serper API Key (Google Search) - https://serper.dev
SERPER_API_KEY=your-serper-api-key
# TAVILY API Key
TAVILY_API_KEY=your-tavily-api-key

View File

@ -0,0 +1,3 @@
from .tools import web_search_tool
__all__ = ["web_search_tool"]

View File

@ -0,0 +1,95 @@
"""
Web Search Tool - Search the web using Serper (Google Search API).
Serper provides real-time Google Search results via a JSON API.
An API key is required. Sign up at https://serper.dev to get one.
"""
import json
import logging
import os
import httpx
from langchain.tools import tool
from deerflow.config import get_app_config
logger = logging.getLogger(__name__)
_SERPER_ENDPOINT = "https://google.serper.dev/search"
_api_key_warned = False
def _get_api_key() -> str | None:
config = get_app_config().get_tool_config("web_search")
if config is not None:
api_key = config.model_extra.get("api_key")
if isinstance(api_key, str) and api_key.strip():
return api_key
return os.getenv("SERPER_API_KEY")
@tool("web_search", parse_docstring=True)
def web_search_tool(query: str, max_results: int = 5) -> str:
"""Search the web for information using Google Search via Serper.
Args:
query: Search keywords describing what you want to find. Be specific for better results.
max_results: Maximum number of search results to return. Default is 5.
"""
global _api_key_warned
config = get_app_config().get_tool_config("web_search")
if config is not None and "max_results" in config.model_extra:
max_results = config.model_extra.get("max_results", max_results)
api_key = _get_api_key()
if not api_key:
if not _api_key_warned:
_api_key_warned = True
logger.warning("Serper API key is not set. Set SERPER_API_KEY in your environment or provide api_key in config.yaml. Sign up at https://serper.dev")
return json.dumps(
{"error": "SERPER_API_KEY is not configured", "query": query},
ensure_ascii=False,
)
headers = {
"X-API-KEY": api_key,
"Content-Type": "application/json",
}
payload = {"q": query, "num": max_results}
try:
with httpx.Client(timeout=30) as client:
response = client.post(_SERPER_ENDPOINT, headers=headers, json=payload)
response.raise_for_status()
data = response.json()
except httpx.HTTPStatusError as e:
logger.error(f"Serper API returned HTTP {e.response.status_code}: {e.response.text}")
return json.dumps(
{"error": f"Serper API error: HTTP {e.response.status_code}", "query": query},
ensure_ascii=False,
)
except Exception as e:
logger.error(f"Serper search failed: {type(e).__name__}: {e}")
return json.dumps({"error": str(e), "query": query}, ensure_ascii=False)
organic = data.get("organic", [])
if not organic:
return json.dumps({"error": "No results found", "query": query}, ensure_ascii=False)
normalized_results = [
{
"title": r.get("title", ""),
"url": r.get("link", ""),
"content": r.get("snippet", ""),
}
for r in organic[:max_results]
]
output = {
"query": query,
"total_results": len(normalized_results),
"results": normalized_results,
}
return json.dumps(output, indent=2, ensure_ascii=False)

View File

@ -0,0 +1,308 @@
"""Unit tests for the Serper community web search tool."""
import json
from unittest.mock import MagicMock, patch
import httpx
import pytest
@pytest.fixture(autouse=True)
def reset_api_key_warned():
"""Reset the module-level warning flag before each test."""
import deerflow.community.serper.tools as serper_mod
serper_mod._api_key_warned = False
yield
serper_mod._api_key_warned = False
@pytest.fixture
def mock_config_with_key():
with patch("deerflow.community.serper.tools.get_app_config") as mock:
tool_config = MagicMock()
tool_config.model_extra = {"api_key": "test-serper-key", "max_results": 5}
mock.return_value.get_tool_config.return_value = tool_config
yield mock
@pytest.fixture
def mock_config_no_key():
with patch("deerflow.community.serper.tools.get_app_config") as mock:
tool_config = MagicMock()
tool_config.model_extra = {}
mock.return_value.get_tool_config.return_value = tool_config
yield mock
def _make_serper_response(organic: list) -> MagicMock:
mock_resp = MagicMock()
mock_resp.json.return_value = {"organic": organic}
mock_resp.raise_for_status = MagicMock()
return mock_resp
class TestGetApiKey:
def test_returns_config_key_when_present(self):
with patch("deerflow.community.serper.tools.get_app_config") as mock:
tool_config = MagicMock()
tool_config.model_extra = {"api_key": "from-config"}
mock.return_value.get_tool_config.return_value = tool_config
from deerflow.community.serper.tools import _get_api_key
assert _get_api_key() == "from-config"
def test_falls_back_to_env_when_config_key_empty(self):
with patch("deerflow.community.serper.tools.get_app_config") as mock:
tool_config = MagicMock()
tool_config.model_extra = {"api_key": ""}
mock.return_value.get_tool_config.return_value = tool_config
with patch.dict("os.environ", {"SERPER_API_KEY": "env-key"}):
from deerflow.community.serper.tools import _get_api_key
assert _get_api_key() == "env-key"
def test_falls_back_to_env_when_config_key_whitespace(self):
with patch("deerflow.community.serper.tools.get_app_config") as mock:
tool_config = MagicMock()
tool_config.model_extra = {"api_key": " "}
mock.return_value.get_tool_config.return_value = tool_config
with patch.dict("os.environ", {"SERPER_API_KEY": "env-key"}):
from deerflow.community.serper.tools import _get_api_key
assert _get_api_key() == "env-key"
def test_falls_back_to_env_when_config_key_null(self):
with patch("deerflow.community.serper.tools.get_app_config") as mock:
tool_config = MagicMock()
tool_config.model_extra = {"api_key": None}
mock.return_value.get_tool_config.return_value = tool_config
with patch.dict("os.environ", {"SERPER_API_KEY": "env-key"}):
from deerflow.community.serper.tools import _get_api_key
assert _get_api_key() == "env-key"
def test_falls_back_to_env_when_no_config(self):
with patch("deerflow.community.serper.tools.get_app_config") as mock:
mock.return_value.get_tool_config.return_value = None
with patch.dict("os.environ", {"SERPER_API_KEY": "env-only"}):
from deerflow.community.serper.tools import _get_api_key
assert _get_api_key() == "env-only"
def test_returns_none_when_no_key_anywhere(self):
with patch("deerflow.community.serper.tools.get_app_config") as mock:
mock.return_value.get_tool_config.return_value = None
with patch.dict("os.environ", {}, clear=True):
import os
os.environ.pop("SERPER_API_KEY", None)
from deerflow.community.serper.tools import _get_api_key
assert _get_api_key() is None
class TestWebSearchTool:
def test_basic_search_returns_normalized_results(self, mock_config_with_key):
organic = [
{"title": "Result 1", "link": "https://example.com/1", "snippet": "Snippet 1"},
{"title": "Result 2", "link": "https://example.com/2", "snippet": "Snippet 2"},
]
mock_resp = _make_serper_response(organic)
with patch("deerflow.community.serper.tools.httpx.Client") as mock_client_cls:
mock_client_cls.return_value.__enter__.return_value.post.return_value = mock_resp
from deerflow.community.serper.tools import web_search_tool
result = web_search_tool.invoke({"query": "python tutorial"})
parsed = json.loads(result)
assert parsed["query"] == "python tutorial"
assert parsed["total_results"] == 2
assert parsed["results"][0]["title"] == "Result 1"
assert parsed["results"][0]["url"] == "https://example.com/1"
assert parsed["results"][0]["content"] == "Snippet 1"
def test_respects_max_results_from_config(self, mock_config_with_key):
mock_config_with_key.return_value.get_tool_config.return_value.model_extra = {
"api_key": "test-key",
"max_results": 3,
}
organic = [{"title": f"R{i}", "link": f"https://x.com/{i}", "snippet": f"S{i}"} for i in range(10)]
mock_resp = _make_serper_response(organic)
with patch("deerflow.community.serper.tools.httpx.Client") as mock_client_cls:
mock_client_cls.return_value.__enter__.return_value.post.return_value = mock_resp
from deerflow.community.serper.tools import web_search_tool
result = web_search_tool.invoke({"query": "test"})
parsed = json.loads(result)
assert parsed["total_results"] == 3
assert len(parsed["results"]) == 3
def test_max_results_parameter_accepted(self, mock_config_no_key):
"""Tool accepts max_results as a call parameter when config does not override it."""
organic = [{"title": f"R{i}", "link": f"https://x.com/{i}", "snippet": f"S{i}"} for i in range(10)]
mock_resp = _make_serper_response(organic)
with patch.dict("os.environ", {"SERPER_API_KEY": "env-key"}):
with patch("deerflow.community.serper.tools.httpx.Client") as mock_client_cls:
mock_client_cls.return_value.__enter__.return_value.post.return_value = mock_resp
from deerflow.community.serper.tools import web_search_tool
result = web_search_tool.invoke({"query": "test", "max_results": 2})
parsed = json.loads(result)
assert parsed["total_results"] == 2
def test_config_max_results_overrides_parameter(self):
"""Config max_results overrides the parameter passed at call time, matching ddg_search behaviour."""
with patch("deerflow.community.serper.tools.get_app_config") as mock:
tool_config = MagicMock()
tool_config.model_extra = {"api_key": "test-key", "max_results": 3}
mock.return_value.get_tool_config.return_value = tool_config
organic = [{"title": f"R{i}", "link": f"https://x.com/{i}", "snippet": f"S{i}"} for i in range(10)]
mock_resp = _make_serper_response(organic)
with patch("deerflow.community.serper.tools.httpx.Client") as mock_client_cls:
mock_client_cls.return_value.__enter__.return_value.post.return_value = mock_resp
from deerflow.community.serper.tools import web_search_tool
result = web_search_tool.invoke({"query": "test", "max_results": 8})
parsed = json.loads(result)
assert parsed["total_results"] == 3
def test_empty_organic_returns_error_json(self, mock_config_with_key):
"""Empty organic list returns structured error, matching ddg_search convention."""
mock_resp = _make_serper_response([])
with patch("deerflow.community.serper.tools.httpx.Client") as mock_client_cls:
mock_client_cls.return_value.__enter__.return_value.post.return_value = mock_resp
from deerflow.community.serper.tools import web_search_tool
result = web_search_tool.invoke({"query": "no results"})
parsed = json.loads(result)
assert "error" in parsed
assert parsed["error"] == "No results found"
assert parsed["query"] == "no results"
def test_missing_api_key_returns_error_json(self, mock_config_no_key):
with patch.dict("os.environ", {}, clear=True):
import os
os.environ.pop("SERPER_API_KEY", None)
from deerflow.community.serper.tools import web_search_tool
result = web_search_tool.invoke({"query": "test"})
parsed = json.loads(result)
assert "error" in parsed
assert "SERPER_API_KEY" in parsed["error"]
def test_missing_api_key_logs_warning_once(self, mock_config_no_key, caplog):
import logging
with patch.dict("os.environ", {}, clear=True):
import os
os.environ.pop("SERPER_API_KEY", None)
from deerflow.community.serper.tools import web_search_tool
with caplog.at_level(logging.WARNING, logger="deerflow.community.serper.tools"):
web_search_tool.invoke({"query": "q1"})
web_search_tool.invoke({"query": "q2"})
warnings = [r for r in caplog.records if r.levelno == logging.WARNING]
assert len(warnings) == 1
def test_http_error_returns_structured_error(self, mock_config_with_key):
mock_error_response = MagicMock()
mock_error_response.status_code = 403
mock_error_response.text = "Forbidden"
with patch("deerflow.community.serper.tools.httpx.Client") as mock_client_cls:
mock_client_cls.return_value.__enter__.return_value.post.side_effect = httpx.HTTPStatusError("403", request=MagicMock(), response=mock_error_response)
from deerflow.community.serper.tools import web_search_tool
result = web_search_tool.invoke({"query": "test"})
parsed = json.loads(result)
assert "error" in parsed
assert "403" in parsed["error"]
def test_network_exception_returns_error_json(self, mock_config_with_key):
with patch("deerflow.community.serper.tools.httpx.Client") as mock_client_cls:
mock_client_cls.return_value.__enter__.return_value.post.side_effect = Exception("timeout")
from deerflow.community.serper.tools import web_search_tool
result = web_search_tool.invoke({"query": "test"})
parsed = json.loads(result)
assert "error" in parsed
def test_sends_correct_headers_and_payload(self, mock_config_with_key):
organic = [{"title": "T", "link": "https://x.com", "snippet": "S"}]
mock_resp = _make_serper_response(organic)
with patch("deerflow.community.serper.tools.httpx.Client") as mock_client_cls:
mock_post = mock_client_cls.return_value.__enter__.return_value.post
mock_post.return_value = mock_resp
from deerflow.community.serper.tools import web_search_tool
web_search_tool.invoke({"query": "hello world"})
call_kwargs = mock_post.call_args
headers = call_kwargs.kwargs["headers"]
payload = call_kwargs.kwargs["json"]
assert headers["X-API-KEY"] == "test-serper-key"
assert payload["q"] == "hello world"
assert payload["num"] == 5
def test_uses_env_key_when_config_absent(self):
with patch("deerflow.community.serper.tools.get_app_config") as mock:
mock.return_value.get_tool_config.return_value = None
with patch.dict("os.environ", {"SERPER_API_KEY": "env-only-key"}):
organic = [{"title": "T", "link": "https://x.com", "snippet": "S"}]
mock_resp = _make_serper_response(organic)
with patch("deerflow.community.serper.tools.httpx.Client") as mock_client_cls:
mock_post = mock_client_cls.return_value.__enter__.return_value.post
mock_post.return_value = mock_resp
from deerflow.community.serper.tools import web_search_tool
web_search_tool.invoke({"query": "env key test"})
headers = mock_post.call_args.kwargs["headers"]
assert headers["X-API-KEY"] == "env-only-key"
def test_partial_fields_in_organic_result(self, mock_config_with_key):
"""Missing title/link/snippet should default to empty string."""
organic = [{}]
mock_resp = _make_serper_response(organic)
with patch("deerflow.community.serper.tools.httpx.Client") as mock_client_cls:
mock_client_cls.return_value.__enter__.return_value.post.return_value = mock_resp
from deerflow.community.serper.tools import web_search_tool
result = web_search_tool.invoke({"query": "test"})
parsed = json.loads(result)
assert parsed["results"][0] == {"title": "", "url": "", "content": ""}

View File

@ -373,6 +373,16 @@ tools:
use: deerflow.community.ddg_search.tools:web_search_tool
max_results: 5
# Web search tool (uses Serper - Google Search API, requires SERPER_API_KEY)
# Serper provides real-time Google Search results. Sign up at https://serper.dev
# Note: set SERPER_API_KEY in your environment before starting the app, or
# uncomment and fill in api_key below (the $VAR syntax is resolved at startup).
# - name: web_search
# group: web
# use: deerflow.community.serper.tools:web_search_tool
# max_results: 5
# # api_key: $SERPER_API_KEY # Optional if SERPER_API_KEY env var is set
# Web search tool (requires Tavily API key)
# - name: web_search
# group: web