chore(backend): 补充核心流程注释与待办

backend/app/agent/permissions.py

@@ -1,3 +1,5 @@
+"""Agent 工具权限策略与一次性确认票据。"""
+
 import asyncio
 from dataclasses import dataclass
 from enum import Enum
@@ -51,6 +53,7 @@ class PermissionPolicy:
     def mode_for(self, permission: str | None) -> PermissionMode:
         if permission is None:
             return PermissionMode.allow
+        # 未登记权限一律拒绝，防止扩展通过拼写错误或新权限绕过策略。
         return self._rules.get(permission, PermissionMode.deny)
 
 
@@ -63,6 +66,8 @@ class PermissionTicket:
 
 
 class PermissionManager:
+    """管理当前进程内的确认请求与会话级授权。"""
+
     def __init__(self, policy: PermissionPolicy) -> None:
         self.policy = policy
         self._pending: dict[tuple[str, str], PermissionTicket] = {}
@@ -94,6 +99,7 @@ class PermissionManager:
         if ticket is None or ticket.future.done():
             return False
         if decision == "allow_session":
+            # 会话授权只存在于进程内，应用重启后按默认策略重新确认。
             self._session_grants.add(ticket.permission)
         ticket.future.set_result(decision)
         return True

backend/app/agent/runtime.py

@@ -1,3 +1,5 @@
+"""Agent 运行时：负责模型轮次、工具调用、权限确认与事件发布。"""
+
 from __future__ import annotations
 
 import asyncio
@@ -50,6 +52,8 @@ MAX_TOOL_CALLS_PER_TURN = 50
 
 @dataclass(slots=True)
 class RunRecord:
+    """单次运行的可变上下文，仅由 AgentRuntime 持有。"""
+
     run: AgentRun
     request: AgentRunCreateRequest
     skill_config: AgentConfiguration | None = None
@@ -60,6 +64,8 @@ class RunRecord:
 
 
 class AgentRuntime:
+    """进程内 Agent 编排器；对外返回深拷贝，避免调用方修改运行状态。"""
+
     def __init__(
         self,
         providers: ProviderRegistry,
@@ -98,6 +104,7 @@ class AgentRuntime:
         )
         allowed_tools = list(request.allowed_tools)
         if skill_config is not None:
+            # 同时指定 Skill 与工具白名单时取交集，避免 Skill 扩大调用权限。
             allowed_tools = (
                 [name for name in skill_config.allowed_tools if name in allowed_tools]
                 if allowed_tools
@@ -142,6 +149,8 @@ class AgentRuntime:
 
     async def events(self, run_id: str) -> AsyncIterator[AgentEvent]:
         record = self._get_record(run_id)
+        # 先回放快照再订阅实时事件，使晚加入的 SSE 客户端也能恢复界面状态。
+        # TODO(agent): 持久化事件并支持 Last-Event-ID，进程重启后仍可续传。
         queue: asyncio.Queue[AgentEvent] = asyncio.Queue()
         record.subscribers.add(queue)
         history = [event.model_copy(deep=True) for event in record.events]
@@ -243,6 +252,7 @@ class AgentRuntime:
                 messages.append(
                     Message(role=MessageRole.assistant, content=turn.text or "", tool_calls=calls)
                 )
+                # 工具可以并发执行，但结果按模型原始调用顺序写回上下文，保证轮次可复现。
                 semaphore = asyncio.Semaphore(record.request.max_concurrent_tools)
 
                 async def execute(call: ToolCall) -> ToolResult:
@@ -313,6 +323,7 @@ class AgentRuntime:
         if mode == PermissionMode.deny:
             result = self._permission_denied(call)
         elif mode == PermissionMode.confirm and permission:
+            # 运行状态必须在等待期间可见，前端才能展示并处理权限确认卡片。
             ticket = self.permissions.create_ticket(record.run.run_id, permission)
             record.run.status = AgentRunStatus.waiting_permission
             self._publish(
@@ -408,6 +419,7 @@ class AgentRuntime:
             timestamp=datetime.now(timezone.utc),
         )
         record.events.append(event)
+        # 内存事件只保留最近窗口；完整审计轨迹应由后续持久化层承担。
         if len(record.events) > MAX_EVENTS_PER_RUN:
             del record.events[: len(record.events) - MAX_EVENTS_PER_RUN]
         for queue in record.subscribers:
@@ -448,6 +460,7 @@ class AgentRuntime:
             raise AgentRunNotFoundError(run_id) from exc
 
     def _prune_records(self) -> None:
+        # 只清理终态记录，绝不为了容量取消仍在执行或等待授权的任务。
         overflow = len(self._records) - MAX_RUN_RECORDS + 1
         if overflow <= 0:
             return

backend/app/agent/tools.py

@@ -1,3 +1,5 @@
+"""Agent 工具注册与执行边界。"""
+
 import inspect
 from dataclasses import dataclass
 from time import perf_counter
@@ -29,6 +31,8 @@ class ToolNotFoundError(LookupError):
 
 
 class ToolRegistry:
+    """统一校验工具入参并隔离执行异常，避免单个工具击穿 Agent 主循环。"""
+
     def __init__(self) -> None:
         self._tools: dict[str, RegisteredTool] = {}
 
@@ -80,6 +84,7 @@ class ToolRegistry:
             )
 
         try:
+            # JSON Schema 约束模型可见的协议，Pydantic 再完成运行时类型转换。
             Draft202012Validator(registered.definition.parameters).validate(call.arguments)
             arguments = registered.arguments_model.model_validate(call.arguments)
         except (ValidationError, JsonSchemaValidationError) as exc:
@@ -103,7 +108,7 @@ class ToolRegistry:
                 output=output,
                 duration_ms=round((perf_counter() - started) * 1000),
             )
-        except Exception as exc:  # Tool failures are isolated from the Agent loop.
+        except Exception as exc:  # 工具失败转换成结构化结果，由模型决定是否降级或重试。
             return ToolResult(
                 tool_call_id=call.tool_call_id,
                 name=call.name,