Update readme with latest options

README.md

@@ -21,7 +21,6 @@ https://github.com/user-attachments/assets/f3e60fff-8680-4dd9-b08e-fa7db655a705
         "airflow-mcp-server"
       ],
       "env": {
-        "OPENAPI_SPEC": "<path_to_spec.yaml>",
         "AIRFLOW_BASE_URL": "http://<host:port>/api/v1",
         "AUTH_TOKEN": "<base64_encoded_username_password>"
       }
@@ -30,12 +29,13 @@ https://github.com/user-attachments/assets/f3e60fff-8680-4dd9-b08e-fa7db655a705
 }
 ```
 
-> You can download the openapi spec from [Airflow REST API](https://airflow.apache.org/docs/apache-airflow/stable/stable-rest-api-ref.html)
 
 # Scope
 
 2 different streams in which Airflow MCP Server can be used:
 - Adding Airflow to AI  (_complete access to an Airflow deployment_)
   - This will enable AI to be able to write DAGs and just do things in a schedule on its own.
+  - Use command `airflow-mcp-server` or `airflow-mcp-server --unsafe`.
 - Adding AI to Airflow (_read-only access using Airflow Plugin_)
   - This stream can enable Users to be able to get a better understanding about their deployment. Specially in cases where teams have hundreds, if not thousands of dags.
+  - Use command `airflow-mcp-server --safe`.