feat: migrate web search tool to support Tavily alongside DuckDuckGo

.env.example

@@ -19,6 +19,7 @@ DATABASE_URL=postgresql://postgres:postgres@localhost:5432/openworkers
 ANTHROPIC_API_KEY=your_anthropic_api_key_here
 OPENAI_API_KEY=your_openai_api_key_here
 DEEPSEEK_API_KEY=your_deepseek_api_key_here
+TAVILY_API_KEY=your_tavily_api_key_here
 
 # ─── Per-Mode Provider + Model ──────────────────────────────────
 # Three modes control which provider AND model each agent uses:

core/config.py

@@ -59,6 +59,7 @@ class Settings(BaseSettings):
     anthropic_api_key: str = ""
     openai_api_key: str = ""
     deepseek_api_key: str = ""
+    tavily_api_key: str = ""
 
     # ── budget guard ────────────────────────────────────────────────────
     max_budget_usd: Optional[float] = None

pyproject.toml

@@ -14,6 +14,7 @@ dependencies = [
     "anthropic>=0.25.0",
     "openai>=1.14.0",
     "duckduckgo-search>=5.2.0",
+    "tavily-python>=0.5.0",
     "pymupdf>=1.24.0",
     "httpx>=0.27.0",
     "tenacity>=9.0.0",

tools/mcp/engine.py

@@ -5,7 +5,9 @@
 from typing import Any
 
 from duckduckgo_search import DDGS
+from tavily import TavilyClient
 
+from core.config import get_settings
 from tools.cache import SearchCache, get_default_cache
 
 
@@ -93,6 +95,28 @@ async def execute_impl(self, params: dict[str, Any]) -> dict[str, Any]:
         if not query:
             return {"results": []}
 
+        tavily_api_key = get_settings().tavily_api_key
+        if tavily_api_key:
+            try:
+                return await self._tavily_search(query, tavily_api_key)
+            except Exception as e:
+                logging.warning(f"Tavily search failed, falling back to DDGS: {e}")
+
+        return await self._ddgs_search(query)
+
+    async def _tavily_search(self, query: str, api_key: str) -> dict[str, Any]:
+        def _search():
+            client = TavilyClient(api_key=api_key)
+            return client.search(query=query, max_results=5)
+
+        response = await asyncio.to_thread(_search)
+        formatted_results = [
+            f"{r.get('title', '')} - {r.get('content', '')} ({r.get('url', '')})"
+            for r in response.get("results", [])
+        ]
+        return {"results": formatted_results}
+
+    async def _ddgs_search(self, query: str) -> dict[str, Any]:
         def _search():
             with DDGS() as ddgs:
                 return list(ddgs.text(query, max_results=5))