feat: add Tavily search tool as configurable alternative

.env.template

@@ -1,3 +1,6 @@
 
 # Must have for minimal agent get started
-OPENROUTER_API_KEY="sk-or-v1-xxxxxxxx99b"
+OPENROUTER_API_KEY="sk-or-v1-xxxxxxxx99b"
+
+# Required when using tool-searching-tavily tool config
+TAVILY_API_KEY=

config/tool/tool-searching-tavily.yaml

@@ -0,0 +1,8 @@
+name: "tool-searching-tavily"
+tool_command: "python"
+args:
+  - "-m"
+  - "src.tool.mcp_servers.searching_tavily_mcp_server"
+env:
+# Tavily API key - this value will be loaded from the .env file at runtime
+  TAVILY_API_KEY: "${oc.env:TAVILY_API_KEY}"

pyproject.toml

@@ -30,7 +30,8 @@ dependencies = [
     "requests>=2.32.4",
     "filetype>=1.2.0",
     "Pillow",
-    "pyzmq"
+    "pyzmq",
+    "tavily-python>=0.5.0"
 ]
 
 [build-system]

src/tool/mcp_servers/searching_tavily_mcp_server.py

@@ -0,0 +1,102 @@
+# SPDX-FileCopyrightText: 2025 MiromindAI
+#
+# SPDX-License-Identifier: Apache-2.0
+
+import os
+import json
+from fastmcp import FastMCP
+from tavily import TavilyClient
+from src.logging.logger import setup_mcp_logging
+
+
+TAVILY_API_KEY = os.environ.get("TAVILY_API_KEY", "")
+
+# Initialize FastMCP server
+setup_mcp_logging(tool_name=os.path.basename(__file__))
+mcp = FastMCP("searching-tavily-mcp-server")
+
+
+@mcp.tool()
+async def tavily_search(
+    query: str,
+    max_results: int = 10,
+    search_depth: str = "advanced",
+    topic: str = "general",
+    time_range: str = None,
+    include_domains: list[str] = None,
+    exclude_domains: list[str] = None,
+) -> str:
+    """Perform web searches via the Tavily API and retrieve relevant results.
+
+    Args:
+        query: Search query string (keep under 400 characters for best results).
+        max_results: The number of results to return (default: 10).
+        search_depth: Search depth - 'basic' for quick results or 'advanced' for highest relevance (default: 'advanced').
+        topic: Search topic category - 'general', 'news', or 'finance' (default: 'general').
+        time_range: Time filter for results - 'day', 'week', 'month', 'year', or None for no filter.
+        include_domains: List of domains to restrict search to (e.g. ['example.com']).
+        exclude_domains: List of domains to exclude from search.
+
+    Returns:
+        The search results as a JSON string.
+    """
+    if not TAVILY_API_KEY:
+        return "[ERROR]: TAVILY_API_KEY is not set, tavily_search tool is not available."
+
+    try:
+        client = TavilyClient(api_key=TAVILY_API_KEY)
+        kwargs = {
+            "query": query,
+            "max_results": max_results,
+            "search_depth": search_depth,
+            "topic": topic,
+        }
+        if time_range:
+            kwargs["time_range"] = time_range
+        if include_domains:
+            kwargs["include_domains"] = include_domains
+        if exclude_domains:
+            kwargs["exclude_domains"] = exclude_domains
+
+        response = client.search(**kwargs)
+        return json.dumps(response, ensure_ascii=False, indent=2)
+
+    except Exception as e:
+        return f"[ERROR]: tavily_search tool execution failed: {str(e)}"
+
+
+@mcp.tool()
+async def scrape_website(url: str) -> str:
+    """Scrape a website for its content using Tavily's extract API.
+
+    This tool extracts the main content from a given URL. It can be used to get
+    webpage content, article text, documentation pages, etc.
+
+    Args:
+        url: The URL of the website to scrape.
+
+    Returns:
+        The scraped website content.
+    """
+    if not TAVILY_API_KEY:
+        return "[ERROR]: TAVILY_API_KEY is not set, scrape_website tool is not available."
+
+    try:
+        client = TavilyClient(api_key=TAVILY_API_KEY)
+        response = client.extract(urls=[url])
+
+        if response.get("results"):
+            result = response["results"][0]
+            return result.get("raw_content", result.get("text", ""))
+        elif response.get("failed_results"):
+            failed = response["failed_results"][0]
+            return f"[ERROR]: Failed to extract content from {url}: {failed.get('error', 'Unknown error')}"
+        else:
+            return f"[ERROR]: No content extracted from {url}"
+
+    except Exception as e:
+        return f"[ERROR]: scrape_website tool execution failed: {str(e)}"
+
+
+if __name__ == "__main__":
+    mcp.run(transport="stdio", show_banner=False)