fix(search): coerce DDG result limits from environment variables (#5792)

* fix(search): coerce DDG result limits from environment variables

* fix(search): fall back for invalid DDG result limits

---------

Co-authored-by: kwsYegar <331533331+kwsYegar@users.noreply.github.com>

README.md

@@ -1172,6 +1172,8 @@ Public-skill CI waivers are exact, expiring exceptions in `.github/skill-review-
 
 Tools follow the same philosophy. DeerFlow comes with a core toolset — web search, web fetch, rendered web capture, file operations, bash execution — and supports custom tools via MCP servers and Python functions. The bundled DDG, Brave, Tavily, and SearXNG search providers accept an optional `time_range` of `day`, `week`, `month`, or `year`; omitting it preserves existing search behavior. For DDG recency searches, DeerFlow excludes DDGS backends that ignore time limits. Swap anything. Add anything.
 
+For DDG search, `max_results` in `config.yaml` can be a positive integer or an environment-variable reference such as `max_results: $DDG_MAX_RESULTS` with `DDG_MAX_RESULTS=5`. The configured value takes precedence over the tool call's `max_results` argument. Invalid values (for example, `abc`, an empty string, or `3.5`), zero, and negative counts produce a warning and fall back to the default of 5 results.
+
 Stdio MCP servers can set `cwd` in `extensions_config.json` when their entrypoint
 or data files depend on a specific working directory. The setting applies to
 both discovery and tool calls; see [MCP configuration](backend/docs/MCP_SERVER.md#stdio-working-directory).

backend/packages/harness/deerflow/community/ddg_search/tools.py

@@ -13,6 +13,7 @@ from deerflow.config import get_app_config
 logger = logging.getLogger(__name__)
 
 DEFAULT_BACKEND = "auto"
+DEFAULT_MAX_RESULTS = 5
 DEFAULT_REGION = "wt-wt"
 DEFAULT_SAFESEARCH = "moderate"
 DEFAULT_WIKIPEDIA_REGION = "us-en"
@@ -30,6 +31,18 @@ WIKIPEDIA_LANGUAGE_ALIASES = {
 }
 
 
+def _coerce_max_results(value: object) -> int:
+    """Normalize config/parameter values before passing them to DDGS."""
+    try:
+        count = int(value)  # type: ignore[call-overload]
+    except (TypeError, ValueError, OverflowError):
+        count = 0
+    if count <= 0:
+        logger.warning("Invalid DDG Search max_results=%r; using default %s", value, DEFAULT_MAX_RESULTS)
+        return DEFAULT_MAX_RESULTS
+    return count
+
+
 def _normalize_backend(backend: str | list[str] | tuple[str, ...] | None) -> str:
     if backend is None:
         return DEFAULT_BACKEND
@@ -105,7 +118,7 @@ def _resolve_ddgs_region(query: str, region: str | None, backend: str | list[str
 
 def _search_text(
     query: str,
-    max_results: int = 5,
+    max_results: int = DEFAULT_MAX_RESULTS,
     region: str | None = DEFAULT_REGION,
     safesearch: str | None = DEFAULT_SAFESEARCH,
     backend: str | list[str] | tuple[str, ...] | None = DEFAULT_BACKEND,
@@ -156,7 +169,7 @@ def _search_text(
 @tool("web_search", parse_docstring=True)
 def web_search_tool(
     query: str,
-    max_results: int = 5,
+    max_results: int = DEFAULT_MAX_RESULTS,
     time_range: SearchTimeRange | None = None,
 ) -> str:
     """Search the web for information. Use this tool to find current information, news, articles, and facts from the internet.
@@ -180,7 +193,7 @@ def web_search_tool(
 
     results = _search_text(
         query=query,
-        max_results=max_results,
+        max_results=_coerce_max_results(max_results),
         region=region,
         safesearch=safesearch,
         backend=backend,

backend/tests/test_ddg_search_tools.py

@@ -1,6 +1,7 @@
 """Unit tests for the DDGS community web search tool."""
 
 import json
+import logging
 import sys
 from types import SimpleNamespace
 from unittest.mock import MagicMock, patch
@@ -10,6 +11,60 @@ import pytest
 from deerflow.community.ddg_search import tools
 
 
+@pytest.mark.parametrize(
+    ("configured_limit", "env_value", "call_args", "expected_count", "warns"),
+    [
+        pytest.param("$DDG_TEST_MAX_RESULTS", "3", {"max_results": 2}, 3, False, id="environment-variable"),
+        pytest.param("3", "3", {"max_results": 2}, 3, False, id="numeric-string"),
+        pytest.param(3, "3", {"max_results": 2}, 3, False, id="integer-config"),
+        pytest.param(None, "3", {"max_results": 2}, 2, False, id="call-argument"),
+        pytest.param(None, "3", {}, 5, False, id="default"),
+        pytest.param("$DDG_TEST_MAX_RESULTS", "abc", {"max_results": 2}, 5, True, id="invalid-env"),
+        pytest.param("$DDG_TEST_MAX_RESULTS", "", {"max_results": 2}, 5, True, id="empty-env"),
+        pytest.param("$DDG_TEST_MAX_RESULTS", "3.5", {"max_results": 2}, 5, True, id="fractional-env"),
+        pytest.param("$DDG_TEST_MAX_RESULTS", "0", {"max_results": 2}, 5, True, id="zero-env"),
+        pytest.param("$DDG_TEST_MAX_RESULTS", "-2", {"max_results": 2}, 5, True, id="negative-env"),
+        pytest.param(0, "3", {}, 5, True, id="zero-config"),
+        pytest.param(-2, "3", {}, 5, True, id="negative-config"),
+        pytest.param([], "3", {}, 5, True, id="invalid-config-type"),
+        pytest.param(None, "3", {"max_results": 0}, 5, True, id="zero-call-argument"),
+        pytest.param(None, "3", {"max_results": -2}, 5, True, id="negative-call-argument"),
+    ],
+)
+def test_web_search_tool_max_results_with_real_ddgs(monkeypatch, caplog, configured_limit, env_value, call_args, expected_count, warns) -> None:
+    """Exercise SDK count arithmetic and slicing without contacting search engines."""
+    from ddgs.ddgs import DDGS
+    from ddgs.results import TextResult
+
+    from deerflow.config.app_config import AppConfig
+    from deerflow.config.tool_config import ToolConfig
+
+    monkeypatch.setenv("DDG_TEST_MAX_RESULTS", env_value)
+    caplog.set_level(logging.WARNING, logger=tools.__name__)
+    raw_config = {"name": "web_search", "group": "web", "use": "deerflow.community.ddg_search.tools:web_search_tool"}
+    if configured_limit is not None:
+        raw_config["max_results"] = configured_limit
+    tool_config = ToolConfig.model_validate(AppConfig.resolve_env_variables(raw_config))
+    monkeypatch.setattr(tools, "get_app_config", lambda: SimpleNamespace(get_tool_config=lambda name: tool_config))
+
+    rows = [TextResult(title=f"Result {i}", href=f"https://example.com/{i}", body=f"Snippet {i}") for i in range(10)]
+    engine = SimpleNamespace(provider="offline", name="offline", search=MagicMock(return_value=rows))
+    monkeypatch.setattr(DDGS, "_get_network_client", lambda self: None)
+    monkeypatch.setattr(DDGS, "_get_engines", lambda self, category, backend: [engine])
+
+    parsed = json.loads(tools.web_search_tool.invoke({"query": "Result", **call_args}))
+
+    assert "error" not in parsed
+    assert parsed["total_results"] == len(parsed["results"]) == expected_count
+    assert all(result["url"].startswith("https://example.com/") for result in parsed["results"])
+    engine.search.assert_called_once()
+    warnings = [record for record in caplog.records if record.name == tools.__name__ and record.levelno == logging.WARNING]
+    assert len(warnings) == int(warns)
+    if warns:
+        assert "max_results" in warnings[0].getMessage()
+        assert "using default 5" in warnings[0].getMessage()
+
+
 def test_resolve_ddgs_region_maps_worldwide_chinese_query_for_wikipedia() -> None:
     assert tools._resolve_ddgs_region("\u4e16\u754c\u676f\u65b0\u95fb 2026", "wt-wt", "auto") == "cn-zh"