diff --git a/src/phoenix_technologies/gpt_researcher/server.py b/src/phoenix_technologies/gpt_researcher/server.py
index ad14203..fbfa63f 100644
--- a/src/phoenix_technologies/gpt_researcher/server.py
+++ b/src/phoenix_technologies/gpt_researcher/server.py
@@ -18,13 +18,10 @@ from gpt_researcher import GPTResearcher
 load_dotenv()
 
 from utils import (
-    research_store,
     create_success_response,
     handle_exception,
     get_researcher_by_id,
     format_sources_for_response,
-    format_context_with_sources,
-    store_research_results,
     create_research_prompt
 )
 
@@ -54,50 +51,6 @@ class CustomLogsHandler:
         self.logs.append(data)  # Append data to logs
         print(f"MCP Log: {data}")  # For demonstration, print the log
 
-@mcp.resource("research://{topic}")
-async def research_resource(topic: str) -> str:
-    """
-    Provide research context for a given topic directly as a resource.
-
-    This allows LLMs to access web-sourced information without explicit function calls.
-
-    Args:
-        topic: The research topic or query
-
-    Returns:
-        String containing the research context with source information
-    """
-    # Check if we've already researched this topic
-    if topic in research_store:
-        logger.info(f"Returning cached research for topic: {topic}")
-        return research_store[topic]["context"]
-
-    # If not, conduct the research
-    logger.info(f"Conducting new research for resource on topic: {topic}")
-    custom_logs_handler = CustomLogsHandler()
-
-    # Initialize GPT Researcher
-    researcher = GPTResearcher(query=topic, report_type=research_type, websocket=custom_logs_handler)
-
-    try:
-        # Conduct the research
-        await researcher.conduct_research()
-
-        # Get the context and sources
-        context = researcher.get_research_context()
-        sources = researcher.get_research_sources()
-        source_urls = researcher.get_source_urls()
-
-        # Format with sources included
-        formatted_context = format_context_with_sources(topic, context, sources)
-
-        # Store for future use
-        store_research_results(topic, context, sources, source_urls, formatted_context)
-
-        return formatted_context
-    except Exception as e:
-        return f"Error conducting research on '{topic}': {str(e)}"
-
 
 @mcp.tool()
 async def deep_research(query: str) -> Dict[str, Any]:
@@ -132,9 +85,6 @@ async def deep_research(query: str) -> Dict[str, Any]:
         sources = researcher.get_research_sources()
         source_urls = researcher.get_source_urls()
 
-        # Store in the research store for the resource API
-        store_research_results(query, context, sources, source_urls)
-
         return create_success_response({
             "research_id": research_id,
             "query": query,
diff --git a/src/phoenix_technologies/gpt_researcher/utils.py b/src/phoenix_technologies/gpt_researcher/utils.py
index 8ad7d12..283d22d 100644
--- a/src/phoenix_technologies/gpt_researcher/utils.py
+++ b/src/phoenix_technologies/gpt_researcher/utils.py
@@ -11,8 +11,6 @@ from loguru import logger
 # Configure logging for console only (no file logging)
 logger.configure(handlers=[{"sink": sys.stderr, "level": "INFO"}])
 
-# Research store to track ongoing research topics and contexts
-research_store = {}
 
 # API Response Utilities
 def create_error_response(message: str) -> Dict[str, Any]:
@@ -68,44 +66,6 @@ def format_sources_for_response(sources: List[Dict[str, Any]]) -> List[Dict[str,
     ]
 
 
-def format_context_with_sources(topic: str, context: str, sources: List[Dict[str, Any]]) -> str:
-    """
-    Format research context with sources for display.
-
-    Args:
-        topic: Research topic
-        context: Research context
-        sources: List of sources
-
-    Returns:
-        Formatted context string with sources
-    """
-    formatted_context = f"## Research: {topic}\n\n{context}\n\n"
-    formatted_context += "## Sources:\n"
-    for i, source in enumerate(sources):
-        formatted_context += f"{i+1}. {source.get('title', 'Unknown')}: {source.get('url', '')}\n"
-    return formatted_context
-
-
-def store_research_results(topic: str, context: str, sources: List[Dict[str, Any]],
-                           source_urls: List[str], formatted_context: Optional[str] = None):
-    """
-    Store research results in the research store.
-
-    Args:
-        topic: Research topic
-        context: Research context
-        sources: List of sources
-        source_urls: List of source URLs
-        formatted_context: Optional pre-formatted context
-    """
-    research_store[topic] = {
-        "context": formatted_context or context,
-        "sources": sources,
-        "source_urls": source_urls
-    }
-
-
 def create_research_prompt(topic: str, goal: str, report_format: str = "research_report") -> str:
     """
     Create a research query prompt for GPT Researcher.
@@ -125,10 +85,7 @@ def create_research_prompt(topic: str, goal: str, report_format: str = "research
     
     You have two methods to access web-sourced information:
     
-    1. Use the "research://{topic}" resource to directly access context about this topic if it exists
-       or if you want to get straight to the information without tracking a research ID.
-       
-    2. Use the conduct_research tool to perform new research and get a research_id for later use.
+    Use the conduct_research tool to perform new research and get a research_id for later use.
        This tool also returns the context directly in its response, which you can use immediately.
     
     After getting context, you can: