feat(app-instance): 集成Beaver后端并更新配置管理

集成新的Beaver后端服务到应用实例中，替换原有的nanobot实现。

主要变更包括：
- 在Dockerfile和环境配置中添加Beaver相关路径和配置变量
- 更新工作目录结构从.nanobot到.beaver
- 实现Beaver引擎加载器，支持配置文件加载和工具组装
- 添加内置工具如ListDirectoryTool、ReadFileTool、SearchFilesTool
- 更新消息处理流程，支持通道适配器和网关模式
- 重构技能系统，支持显式工具提示和嵌入式检索
- 改进错误处理和生命周期管理

此变更使应用实例能够使用统一的Beaver后端进行AI代理运行时管理。

app-instance/backend/beaver/interfaces/channels/memory.py

@@ -0,0 +1,57 @@
+"""In-memory channel adapter for tests and local gateway embedding."""
+
+from __future__ import annotations
+
+from typing import Any
+
+from beaver.foundation.events import InboundMessage, MessageBus, OutboundMessage
+
+
+class MemoryChannelAdapter:
+    """A local channel that stores outbound messages in memory."""
+
+    def __init__(self, bus: MessageBus, *, name: str = "memory") -> None:
+        self.name = name
+        self.bus = bus
+        self.started = False
+        self.sent_messages: list[OutboundMessage] = []
+
+    async def start(self) -> None:
+        self.started = True
+
+    async def stop(self) -> None:
+        self.started = False
+
+    async def send(self, message: OutboundMessage) -> None:
+        self.sent_messages.append(message)
+
+    async def publish_text(
+        self,
+        content: str,
+        *,
+        session_id: str | None = None,
+        user_id: str | None = None,
+        title: str | None = None,
+        execution_context: str | None = None,
+        model: str | None = None,
+        provider_name: str | None = None,
+        embedding_model: str | None = None,
+        metadata: dict[str, Any] | None = None,
+    ) -> InboundMessage:
+        """Publish a text message from this channel into the shared bus."""
+
+        message = InboundMessage(
+            channel=self.name,
+            content=content,
+            session_id=session_id,
+            user_id=user_id,
+            title=title,
+            execution_context=execution_context,
+            model=model,
+            provider_name=provider_name,
+            embedding_model=embedding_model,
+            metadata=metadata or {},
+        )
+        await self.bus.publish_inbound(message)
+        return message
+