✨ feat!(llm): 重构并升级大语言模型服务为全新 AI 智能体框架 (#2146)

* ✨ feat!(llm): 重构并升级大语言模型服务为全新 AI 智能体框架

- 【重构】将原 services/llm 重构并迁移至全新的 services/ai 架构,提供向下兼容垫片
- 【新增】引入 Agent、Team、Workflow 三大智能体与工作流编排范式
- 【新增】引入基于 RAG 的长期向量记忆与中期槽位记忆系统
- 【新增】引入基于 Docker 的安全代码执行沙箱环境
- 【新增】支持 MCP 协议,允许动态管理和调用 MCP 服务
- 【新增】引入输入输出安全合规护栏与自愈反思机制
- 【优化】重构并优化多厂商 API 适配器 (Gemini, OpenAI, DeepSeek, GLM 等)
- 【优化】优化日志脱敏与 Token 预估机制
- 【移除】移除旧版 llm default 和 llm reset-key 命令,新增 llm mcp 管理命令

* 🔧 chore(deps): 更新项目依赖与配置

- 添加 mcp、jieba 和 aiodocker 依赖到配置文件及 requirements.txt
- 在 pyright 配置中设置 reportMissingImports 为 none
- 调整 .gitignore 中 resources 目录的忽略规则

* ♻️ refactor(tools): 重构工具终止机制并清理知识库日志输出

- 统一使用 `context.state["__end_run__"]` 替代 `EndRunResult` 控制任务结束
- 移除文件系统和向量知识库检索工具中 `ToolResult` 的 `.with_log` 调用
- 调整指令处理器(Directive)的返回值为 `tool_res.output`
- 修复部分类型检查警告并优化联合类型判断语法

* ♻️ refactor(tools): 重构工具副作用指令与控制流熔断机制

- 引入 `DirectivePayload` 及 `ToolResult` 的子类以结构化表达工具副作用
- 移除通过 `context.state` 传递魔术变量的隐式控制流设计
- 重构 `DirectiveManager` 处理器接口,直接在处理器中修改 `AgentState` 并构建 `AgentRunResult`
- 在 `StandardAgentExecutor` 中统一通过 `directive_manager` 调度工具返回的副作用指令
- 补全 `MessageBuilder` 中部分核心方法的文档注释

* 🐛 fix(sandbox): 修复 Docker 沙箱容器状态检测与会话清理逻辑

-【修复】修正 `is_alive` 中直接读取私有属性的问题,改用 `show()` 返回值
-【修复】解决 `execute_code` 中缓存的执行器与当前会话不一致的问题
-【优化】在清理工作区前增加容器存活检测,避免向已死容器发送请求
-【优化】创建容器时增加运行状态校验,若已停止则自动从缓存中移除并重建
-【优化】优化容器销毁和清理逻辑,静默处理容器不存在 (404) 的异常

* 📝 docs(core): 补充核心模块初始化方法的文档注释

* 🚨 auto fix by pre-commit hooks

---------

Co-authored-by: webjoin111 <455457521@qq.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
This commit is contained in:
Rumio
2026-07-03 08:53:56 +08:00
committed by GitHub
co-authored by webjoin111 pre-commit-ci[bot]
parent bdc1374848
commit 80fc5b86a7
223 changed files with 41885 additions and 10006 deletions
+189
View File
@@ -0,0 +1,189 @@
from abc import ABC, abstractmethod
from collections.abc import Awaitable, Callable
from contextlib import AbstractAsyncContextManager
from dataclasses import dataclass
from pathlib import Path
from typing import Any, Protocol, runtime_checkable
from zhenxun.services.ai.sandbox.models import SandboxExecutionResult
class InteractiveTerminalSession(Protocol):
@abstractmethod
async def start(self, cmd: str, env: dict[str, str] | None = None) -> None:
"""启动并挂载终端会话"""
...
@abstractmethod
async def send_input(self, text: str) -> None:
"""向终端发送标准输入"""
...
@abstractmethod
async def read_output(self, timeout: int = 5) -> str:
"""读取当前终端屏幕输出画面"""
...
@abstractmethod
async def interrupt(self) -> None:
"""发送强制中断信号(Ctrl+C)"""
...
@abstractmethod
async def close(self) -> None:
"""释放并关闭终端资源"""
...
@runtime_checkable
class SupportsCommandExecution(Protocol):
async def run_process(
self,
command: str | list[str],
cwd: str | None = None,
timeout: float | None = None,
env: dict[str, str] | None = None,
on_output: Callable[[str, bytes], Awaitable[None]] | None = None,
) -> SandboxExecutionResult:
"""在沙箱内单次执行短命令并获取结果"""
...
@runtime_checkable
class SandboxProcessStream(Protocol):
@abstractmethod
async def read(self) -> "ProcessStreamMessage | None":
"""异步读取下一块输出流数据"""
...
@abstractmethod
async def write(self, data: bytes) -> None:
"""异步写入数据到进程标准输入"""
...
@abstractmethod
async def close(self) -> None:
"""关闭并终止输入输出流"""
...
@runtime_checkable
class SupportsStreamExecution(Protocol):
@abstractmethod
def create_stream_process(
self,
command: str | list[str],
cwd: str | None = None,
env: dict[str, str] | None = None,
) -> AbstractAsyncContextManager[SandboxProcessStream]:
"""创建持久流式后台进程,供长连接通信"""
...
@runtime_checkable
class SupportsInteractivePTY(Protocol):
async def create_pty_session(self) -> InteractiveTerminalSession:
"""创建分配一个真实的伪终端(PTY)交互会话"""
...
@runtime_checkable
class SupportsFileSystem(Protocol):
async def write_raw_file(self, path: str | Path, content: str) -> bool:
"""使用字符串极速覆写文件"""
...
async def read_raw_file(self, path: str | Path) -> str:
"""直接读取文件内容为文本字符串"""
...
async def delete_raw_file(self, path: str | Path) -> bool:
"""直接删除指定物理文件"""
...
async def upload_raw_dir(
self, local_dir_path: str | Path, sandbox_target_path: str | Path
) -> bool:
"""将宿主机本地目录完整打包上传至沙箱"""
...
async def write(self, path: str | Path, data: bytes) -> bool:
"""底层二进制安全写入文件"""
...
async def read(self, path: str | Path) -> bytes:
"""底层二进制安全读取文件"""
...
async def rm(self, path: str | Path, recursive: bool = False) -> bool:
"""执行标准的 rm 删除操作"""
...
async def mkdir(self, path: str | Path, parents: bool = False) -> bool:
"""执行标准的 mkdir 创建目录操作"""
...
@runtime_checkable
class SupportsPortMapping(Protocol):
def get_meta(self, key: str, default: Any = None) -> Any:
"""获取沙箱驱动映射的底层元数据(如分配的随机端口)"""
...
class SandboxChannel(ABC):
@abstractmethod
def get_meta(self, key: str, default: Any = None) -> Any:
"""获取沙箱会话的底层元数据字典"""
...
class StatefulCodeClient(Protocol):
"""有状态代码执行客户端通信协议"""
@abstractmethod
async def execute(
self,
code: str,
timeout: int = 30,
on_output: Callable[[str, bytes], Awaitable[None]] | None = None,
) -> SandboxExecutionResult:
"""执行指定代码并获取结果"""
...
@abstractmethod
async def interrupt(self) -> None:
"""发送强制中断信号(模拟Ctrl+C)"""
...
@abstractmethod
async def close(self) -> None:
"""关闭底层网络及进程连接"""
...
class BaseEngineManager(Protocol):
"""沙箱后台引擎生命周期管理器协议"""
@abstractmethod
async def ensure_started(self, env_vars: dict[str, str] | None = None) -> None:
"""确保后台引擎主服务已在沙箱中成功启动"""
...
@abstractmethod
async def get_client(self, kernel_name: str) -> StatefulCodeClient:
"""分配并获取指定语言内核的通信客户端"""
...
@abstractmethod
async def close(self) -> None:
"""安全关闭引擎并回收所有分配的客户端资源"""
...
@dataclass
class ProcessStreamMessage:
"""统一的进程流消息载体"""
stream_type: int
data: bytes