docs: update codeboarding documentation and generate architecture

## 📚 Documentation Update
This commit contains updated documentation files fetched from the CodeBoarding service and automatically generated architecture documentation.

### 📊 Summary
- Documentation files created/updated: 5
- JSON files created/updated: 6
- Documentation directory: .codeboarding/
- JSON directory: .codeboarding/
- Output format: .mdx
- Repository analyzed: https://github.com/CodeBoarding/mcp-use
- Architecture documentation: docs/development/architecture.mdx (auto-generated)

The generated .mdx files have been automatically appended to the development documentation, and the architecture.mdx file has been generated from the CodeBoarding analysis files.

🤖 This commit was automatically generated by the CodeBoarding documentation update workflow.

Author: actions-user

.codeboarding/Agent_Core.mdx

@@ -19,7 +19,7 @@ graph LR
 
 The `Agent Core` subsystem is the central orchestrator within the `mcp-use` framework, responsible for the agent's decision-making, lifecycle, and overall execution flow. It integrates various components to interact with the external environment and manage its operational state.
 
-### Agent Core [[Expand]](./Agent_Core)
+
 The primary intelligence and orchestration unit. It manages the agent's lifecycle, decision-making processes, and coordinates interactions with other components to execute tasks and communicate with external services.
 
 
@@ -28,7 +28,7 @@ The primary intelligence and orchestration unit. It manages the agent's lifecycl
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/agents/mcpagent.py#L1-L1" target="_blank" rel="noopener noreferrer">`mcp_use.agents.mcpagent.MCPAgent` (1:1)</a>
 
 
-### Base Agent
+
 Provides the foundational structure and common functionalities for all agents within the framework. It serves as an abstract base class that `Agent Core` extends.
 
 
@@ -37,7 +37,7 @@ Provides the foundational structure and common functionalities for all agents wi
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/agents/base.py#L1-L1" target="_blank" rel="noopener noreferrer">`mcp_use.agents.base.BaseAgent` (1:1)</a>
 
 
-### MCP Server Registry/Connector
+
 Manages connections and interactions with various MCP (Multi-Agent Communication Protocol) servers. It acts as an interface for the `Agent Core` to discover and communicate with available services.
 
 
@@ -46,7 +46,7 @@ Manages connections and interactions with various MCP (Multi-Agent Communication
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/managers/server_manager.py#L16-L89" target="_blank" rel="noopener noreferrer">`mcp_use.managers.server_manager.ServerManager` (16:89)</a>
 
 
-### LLM Integration Layer
+
 Facilitates the processing of outputs from Large Language Models (LLMs) and integrates with LangChain-based tools. It acts as an adapter, translating agent requests into LLM-compatible formats and vice-versa.
 
 
@@ -55,7 +55,7 @@ Facilitates the processing of outputs from Large Language Models (LLMs) and inte
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/adapters/langchain_adapter.py#L1-L1" target="_blank" rel="noopener noreferrer">`mcp_use.adapters.langchain_adapter.LangChainAdapter` (1:1)</a>
 
 
-### Telemetry
+
 Responsible for emitting telemetry events, providing insights into the agent's operational status, performance, and interactions. This component supports observability within the `Agent Core` subsystem.
 
 
@@ -64,7 +64,7 @@ Responsible for emitting telemetry events, providing insights into the agent's o
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/telemetry/telemetry.py#L55-L305" target="_blank" rel="noopener noreferrer">`mcp_use.telemetry.telemetry.Telemetry` (55:305)</a>
 
 
-### MCP Session
+
 Manages the state and context of an agent's session. It encapsulates session-specific data and provides mechanisms for maintaining continuity across interactions.
 
 

.codeboarding/Client_Server_Connectivity.mdx

@@ -22,7 +22,7 @@ graph LR
 
 One paragraph explaining the functionality which is represented by this graph. What the main flow is and what is its purpose.
 
-### MCP Client
+
 Serves as the primary client-side interface for initiating and managing interactions with MCP servers. It orchestrates the creation and lifecycle of communication sessions.
 
 
@@ -31,7 +31,7 @@ Serves as the primary client-side interface for initiating and managing interact
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/client.py" target="_blank" rel="noopener noreferrer">`mcp_use.client.MCPClient`</a>
 
 
-### MCP Session Manager
+
 Manages the state, lifecycle, and communication flow of an individual session with an MCP server. It acts as an intermediary, utilizing specific connectors for underlying communication.
 
 
@@ -40,7 +40,7 @@ Manages the state, lifecycle, and communication flow of an individual session wi
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/session.py" target="_blank" rel="noopener noreferrer">`mcp_use.session.MCPSession`</a>
 
 
-### Base Connector Abstraction
+
 Defines the abstract interface for all concrete connection types (e.g., WebSocket, Sandbox). It provides a unified contract for establishing and managing connections, enabling extensibility for new protocols.
 
 
@@ -49,7 +49,7 @@ Defines the abstract interface for all concrete connection types (e.g., WebSocke
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/connectors/base.py" target="_blank" rel="noopener noreferrer">`mcp_use.connectors.base.BaseConnector`</a>
 
 
-### Sandbox Connector
+
 A concrete implementation of the Base Connector Abstraction, specifically designed to handle connections and communication with Sandbox MCP servers.
 
 
@@ -58,7 +58,7 @@ A concrete implementation of the Base Connector Abstraction, specifically design
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/connectors/sandbox.py" target="_blank" rel="noopener noreferrer">`mcp_use.connectors.sandbox.SandboxConnector`</a>
 
 
-### WebSocket Connector
+
 A concrete implementation of the Base Connector Abstraction, responsible for establishing and managing connections using the WebSocket protocol with MCP servers.
 
 
@@ -67,7 +67,7 @@ A concrete implementation of the Base Connector Abstraction, responsible for est
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/connectors/websocket.py" target="_blank" rel="noopener noreferrer">`mcp_use.connectors.websocket.WebSocketConnector`</a>
 
 
-### Connection Task Manager
+
 Manages the asynchronous tasks and operational lifecycle associated with specific connector types, ensuring robust and efficient connection handling.
 
 
@@ -76,7 +76,7 @@ Manages the asynchronous tasks and operational lifecycle associated with specifi
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/task_managers/base.py" target="_blank" rel="noopener noreferrer">`mcp_use.task_managers.base.ConnectionManager`</a>
 
 
-### Server Manager
+
 Manages the lifecycle and state of various MCP servers, providing an interface for clients to connect, disconnect, and retrieve server information. It orchestrates server-side operations and interacts with agents.
 
 
@@ -85,7 +85,7 @@ Manages the lifecycle and state of various MCP servers, providing an interface f
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/managers/server_manager.py#L16-L89" target="_blank" rel="noopener noreferrer">`mcp_use.managers.server_manager.ServerManager` (16:89)</a>
 
 
-### Base Agent Abstraction
+
 Defines the abstract interface for all agent types within the MCP system. It provides a common contract for agents to interact with sessions and execute operations.
 
 

.codeboarding/External_Integration_Layer.json

@@ -17,8 +17,7 @@
           "reference_start_line": 0,
           "reference_end_line": 0
         }
-      ],
-      "can_expand": true
+      ]
     },
     {
       "name": "Base Adapter",
@@ -30,8 +29,7 @@
           "reference_start_line": 0,
           "reference_end_line": 0
         }
-      ],
-      "can_expand": true
+      ]
     },
     {
       "name": "LangChain Adapter",
@@ -43,8 +41,7 @@
           "reference_start_line": 0,
           "reference_end_line": 0
         }
-      ],
-      "can_expand": true
+      ]
     },
     {
       "name": "Connector Management Logic",
@@ -86,8 +83,7 @@
           "reference_start_line": 0,
           "reference_end_line": 0
         }
-      ],
-      "can_expand": true
+      ]
     },
     {
       "name": "Tool/Resource/Prompt Conversion Helpers",
@@ -111,8 +107,7 @@
           "reference_start_line": 0,
           "reference_end_line": 0
         }
-      ],
-      "can_expand": false
+      ]
     }
   ],
   "components_relations": [

.codeboarding/External_Integration_Layer.mdx

@@ -18,7 +18,7 @@ graph LR
 
 Abstract Components Overview
 
-### External Integration Layer [[Expand]](./External_Integration_Layer)
+
 The overarching component responsible for abstracting and managing interactions with external AI frameworks and systems. It acts as the primary interface for `mcp-use` to leverage external functionalities, ensuring loose coupling and adaptability.
 
 
@@ -28,7 +28,7 @@ The overarching component responsible for abstracting and managing interactions
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/adapters/langchain_adapter.py" target="_blank" rel="noopener noreferrer">`mcp_use.adapters.langchain_adapter.LangChainAdapter`</a>
 
 
-### Base Adapter
+
 Defines the abstract interface and contract for all external integrations. It establishes the common methods and properties that any concrete adapter must implement, ensuring a consistent approach to integrating diverse external systems. It also orchestrates the lifecycle of connectors.
 
 
@@ -37,7 +37,7 @@ Defines the abstract interface and contract for all external integrations. It es
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/adapters/base.py" target="_blank" rel="noopener noreferrer">`mcp_use.adapters.base.BaseAdapter`</a>
 
 
-### LangChain Adapter
+
 A concrete implementation of the `BaseAdapter` specifically tailored for integrating with the LangChain framework. It translates `mcp-use`'s internal requests and data structures into LangChain-compatible formats and vice-versa, enabling the use of LangChain's agents, tools, and chains.
 
 
@@ -46,7 +46,7 @@ A concrete implementation of the `BaseAdapter` specifically tailored for integra
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/adapters/langchain_adapter.py" target="_blank" rel="noopener noreferrer">`mcp_use.adapters.langchain_adapter.LangChainAdapter`</a>
 
 
-### Connector Management Logic
+
 An internal helper component responsible for the initialization, configuration, and lifecycle management of connectors to external systems. It ensures that the necessary connections are established and maintained for adapters to function correctly.
 
 
@@ -60,7 +60,7 @@ An internal helper component responsible for the initialization, configuration,
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/session.py" target="_blank" rel="noopener noreferrer">`mcp_use.session.MCPSession`</a>
 
 
-### Tool/Resource/Prompt Conversion Helpers
+
 An internal helper component that performs specific data transformations, such as converting `mcp-use`'s internal tool definitions, resources, or prompt templates into formats compatible with external frameworks (e.g., LangChain's tool schema or prompt templates), and vice-versa.
 
 

.codeboarding/Tool_Abstraction_Management.mdx

@@ -28,7 +28,7 @@ graph LR
 
 Abstract Components Overview
 
-### Agent Core [[Expand]](./Agent_Core)
+
 The central orchestrator of the AI agent framework. It interprets user requests, leverages the LLM Integration Layer for reasoning and decision-making, and utilizes the Tool Abstraction & Management component to dynamically select and execute appropriate tools to achieve its goals. This component embodies the "Agent-based Architecture" pattern.
 
 
@@ -37,7 +37,7 @@ The central orchestrator of the AI agent framework. It interprets user requests,
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/agents/mcpagent.py" target="_blank" rel="noopener noreferrer">`mcp_use.agents.mcpagent.MCPAgent`</a>
 
 
-### LLM Integration Layer
+
 Provides a standardized, abstract interface for interacting with various Large Language Models. It encapsulates the specifics of different LLM APIs, allowing the Agent Core to communicate with them seamlessly for tasks such as natural language understanding, generation, and complex reasoning.
 
 
@@ -46,7 +46,7 @@ Provides a standardized, abstract interface for interacting with various Large L
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/adapters/langchain_adapter.py" target="_blank" rel="noopener noreferrer">`mcp_use.adapters.langchain_adapter.LangChainAdapter`</a>
 
 
-### Tool Abstraction & Management [[Expand]](./Tool_Abstraction_Management)
+
 This component serves as the central Tool/Server Abstraction Layer. It provides a unified, high-level interface for agents to discover, manage, and execute tools exposed by connected MCP servers. It abstracts the complexities of tool interaction, server lifecycle management, and tool execution, aligning with the "Adapter Pattern" and "Client-Server Architecture (Internal)" principles. It is crucial for the framework's extensibility and modularity.
 
 
@@ -61,7 +61,7 @@ This component serves as the central Tool/Server Abstraction Layer. It provides
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/managers/server_manager.py#L16-L89" target="_blank" rel="noopener noreferrer">`mcp_use.managers.server_manager.ServerManager` (16:89)</a>
 
 
-### MCP Server Communication Layer
+
 Responsible for the low-level, concrete communication with external MCP (Multi-Component Platform) servers. It handles network requests, data serialization/deserialization, and error handling specific to server interactions, translating abstract tool execution requests from the Tool Abstraction & Management component into concrete server API calls.
 
 
@@ -73,7 +73,7 @@ Responsible for the low-level, concrete communication with external MCP (Multi-C
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/connectors/websocket.py" target="_blank" rel="noopener noreferrer">`mcp_use.connectors.websocket.WebSocketConnector`</a>
 
 
-### Configuration Management
+
 Manages the loading, parsing, and access of configuration settings for the entire framework. This includes details about registered MCP servers, API keys, and other operational parameters. It ensures that various components can dynamically retrieve necessary settings, supporting the "Configuration-Driven Design" pattern.
 
 
@@ -82,7 +82,7 @@ Manages the loading, parsing, and access of configuration settings for the entir
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/config.py" target="_blank" rel="noopener noreferrer">`mcp_use.config.Config`</a>
 
 
-### Asynchronous Execution Engine
+
 Provides the core asynchronous capabilities for the framework, enabling non-blocking I/O operations and concurrent execution. It is fundamental for efficiently handling long-running tasks such as LLM API calls and external tool executions, leveraging asyncio throughout the system. This aligns with the "Asynchronous Processing" architectural pattern.
 
 

.codeboarding/codeboarding_version.json

@@ -1,4 +1,4 @@
 {
-  "commit_hash": "8cc3acf2df032ca52245da6aa74d793384e44bc6",
+  "commit_hash": "3fae6421f6a7b104109cc034d333c1518a98179a",
   "code_boarding_version": "0.1.0"
 }

.codeboarding/on_boarding.mdx

@@ -29,7 +29,7 @@ This documentation was generated by [CodeBoarding](https://github.com/CodeBoardi
 
 mcp-use is a Python library designed for building and orchestrating AI agents, primarily focusing on interaction with MCP servers. Its architecture is centered around the Agent Core, which serves as the central decision-making and execution unit. This core interacts with the Client & Server Connectivity component to manage connections and sessions with various MCP servers. The Tool Abstraction & Management component provides the agent with a unified way to discover and utilize tools exposed by these connected servers, relying on the connectivity layer to retrieve them. Furthermore, an External Integration Layer allows `mcp-use` to seamlessly adapt and integrate with other AI frameworks like LangChain, ensuring broad interoperability for agent development.
 
-### Agent Core [[Expand]](./Agent_Core)
+
 The central intelligence and orchestration unit, responsible for the agent's decision-making, lifecycle management, and overall execution flow. It leverages other components to interact with the external environment.
 
 
@@ -38,7 +38,7 @@ The central intelligence and orchestration unit, responsible for the agent's dec
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/agents/mcpagent.py#L1-L1" target="_blank" rel="noopener noreferrer">`mcp_use.agents.mcpagent.MCPAgent` (1:1)</a>
 
 
-### Client & Server Connectivity [[Expand]](./Client_Server_Connectivity)
+
 Manages all aspects of establishing, maintaining, and terminating connections with various MCP servers (e.g., WebSocket, Sandbox). It provides the foundational communication layer and handles session management.
 
 
@@ -52,7 +52,7 @@ Manages all aspects of establishing, maintaining, and terminating connections wi
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/task_managers/base.py#L1-L1" target="_blank" rel="noopener noreferrer">`mcp_use.task_managers.base.ConnectionManager` (1:1)</a>
 
 
-### Tool Abstraction & Management [[Expand]](./Tool_Abstraction_Management)
+
 Provides a unified interface for agents to discover, list, search, and execute tools exposed by connected MCP servers. It abstracts the complexities of tool interaction and lifecycle.
 
 
@@ -66,7 +66,7 @@ Provides a unified interface for agents to discover, list, search, and execute t
 - <a href="https://github.com/CodeBoarding/mcp-use/blob/main/mcp_use/managers/tools/use_tool.py#L21-L153" target="_blank" rel="noopener noreferrer">`mcp_use.managers.tools.use_tool.UseToolFromServerTool` (21:153)</a>
 
 
-### External Integration Layer [[Expand]](./External_Integration_Layer)
+
 Acts as an adapter, enabling `mcp-use` to seamlessly integrate with and leverage functionalities from external AI frameworks (e.g., LangChain) or other systems by translating interfaces and data structures.