chore: clean up docs, fix naming, and improve safety

- Remove outdated eval_basic_mcp_use section from README
- Fix naming inconsistency in adk_mcp.py (pydantic -> adk)
- Improve model costs parsing safety with restricted eval
- Fix mermaid diagram string formatting
- Add new evaluation scripts and results directory

Author: andrewginns

README.md

@@ -60,10 +60,6 @@ This project aims to teach:
     - `oai-agent_mcp.py` - Example of using MCP with OpenAI Agents
     - `pydantic_mcp.py` - Example of using MCP with Pydantic-AI
 
-  - **eval_basic_mcp_use/** - Contains evaluation examples for single MCP usage:
-    - `evals_adk_mcp.py` - Evaluation of MCP with Google's ADK
-    - `evals_langchain_mcp.py` - Evaluation of MCP with LangGraph
-    - `evals_pydantic_mcp.py` - Evaluation of MCP with Pydantic-AI
 
 - **[agents_mcp_usage/multi_mcp/](agents_mcp_usage/multi_mcp/)** - Advanced multi-MCP server integration examples
   - **multi_mcp_use/** - Contains examples of using multiple MCP servers simultaneously:

agents_mcp_usage/basic_mcp/basic_mcp_use/adk_mcp.py

@@ -41,20 +41,20 @@ async def main(query: str = "Greet Andrew and give him the current time") -> Non
     # Create the agent
     root_agent = LlmAgent(
         model="gemini-2.5-pro-preview-03-25",
-        name="mcp_pydantic_assistant",
+        name="mcp_adk_assistant",
         tools=tools,
     )
 
     # Set up session
     session_service = InMemorySessionService()
     session = session_service.create_session(
-        app_name="mcp_pydantic_app",
+        app_name="mcp_adk_app",
         user_id="aginns",
     )
 
     # Create the runner
     runner = Runner(
-        app_name="mcp_pydantic_app",
+        app_name="mcp_adk_app",
         agent=root_agent,
         session_service=session_service,
     )

agents_mcp_usage/multi_mcp/eval_multi_mcp/merbench_ui.py

@@ -55,7 +55,11 @@ def load_model_costs(file_path: str) -> tuple[Dict, Dict]:
                 return {}, {}
 
             # Safely evaluate the dictionary strings
-            model_costs_raw = eval(costs_match.group(1), {"float": float})
+            # Handle float('inf') which ast.literal_eval cannot parse
+            costs_string = costs_match.group(1)
+            costs_string = costs_string.replace("float('inf')", "float('inf')")
+            # Use eval with restricted globals for this specific case
+            model_costs_raw = eval(costs_string, {"__builtins__": {}, "float": float})
 
             # Extract friendly names from the inline entries
             friendly_names = {}

agents_mcp_usage/multi_mcp/mermaid_diagrams.py

@@ -198,7 +198,7 @@
 ```
 """
 
-valid_mermaid_diagram = """`
+valid_mermaid_diagram = """
 ```mermaid
 graph LR
     User((User)) --> |"Run script<br>(e.g., pydantic_mcp.py)"| Agent