Files
micro_servers/BotOps/README.md
T
qing 1b9e6f6b3a fix(botops): resolve duplicate AI response rendering in ChatView
Root cause: aibox.py think_and_respond recursively calls itself after tool
execution, and each recursive call's complete callback carries only the
final-round full_response. The old finish_ai logic treated complete as a
full-text alignment signal — when full_text != self._ai_text (which always
happened after recursive tool calls because _ai_text accumulated all chunks
while complete only held the last round), it triggered remove+remount,
causing visible duplicate/flash of AI responses. A secondary issue was
dialogue_ended arriving before complete, which made a second finish_ai call
remount the message a second time.

Changes:
- Remove the full-text replacement branch in finish_ai; chunk streaming is
  now the authoritative content source, complete is a finalization signal only
- Add _ai_finalized flag: start_ai resets it, finish_ai sets it; repeated
  finish_ai calls (dialogue_ended + complete) return early to prevent remount
- append_ai drops late chunks after finalization to avoid dirty appends
- Empty pre-built widget without chunks still gets filled via w.update(full_text)
- Full lost-frame fallback (no widget + full_text) is preserved
- Update BotOps/README.md protocol section to reflect new complete semantics
2026-08-01 02:56:29 +08:00

3.1 KiB
Raw Blame History

BotOps — aibox 智能体 TUI 前端

复刻并增强 aibox/client.py,基于 Textual 的现代化终端 UI,连接 aibox 微服务(WebSocket,默认 ws://localhost:8765)。

相比 client.py 的增强

能力 client.py BotOps
AI 回复渲染 纯文本流式 / 一次性静态渲染 Markdown 增量流式渲染(无闪烁)
终端管理 手搓光标/提示符 Textual 部件框架(多面板布局)
任务事件 混在对话流里 独立侧栏(task_process / task_status / task_error)
自动重连 ✅(已移植) ✅ 指数退避 2s→30s,/disconnect 抑制
快捷键 Ctrl+C 中断 Ctrl+C 中断 / Ctrl+Q 退出 / Ctrl+L 清屏

安装与运行

cd BotOps

# 一键初始化(创建 venv + 安装依赖)
bash _set.sh

# 启动
bash _run.sh                        # 默认 ws://localhost:8765
bash _run.sh ws://your-host:8765    # 指定 aibox 地址

Nuitka 编译为独立二进制

bash _inst.sh        # 产出 ./botops.bin(单文件,无需 Python 环境)
./botops.bin [ws://host:port]

命令

  • 本地命令(不发给服务器):/connect /disconnect /interrupt /clear /status /help /quit
  • 其余 / 命令透传给 aibox 服务器(与 client.py 行为一致)

结构

BotOps/
├── app.py            # Textual App:布局 + 消息路由 + 快捷键
├── ws_client.py      # WebSocket 协议层(自动重连)
├── commands.py       # 斜杠命令注册表
├── _run.sh           # 启动脚本
├── _set.sh           # 初始化脚本(venv + 依赖安装)
├── _inst.sh          # Nuitka 编译脚本(产出 botops.bin)
├── requirements.txt  # Python 依赖
└── widgets/
    ├── __init__.py
    ├── chat_view.py  # 聊天区:AI 回复 Markdown 增量流式渲染
    └── status_bar.py # 底部状态栏(连接状态)

协议说明

与 aibox 服务端协议完全兼容(详见 aibox/client.py 注释):

  • 发送:{"type":"message","content":...} / {"type":"command","command":"/..."} / {"type":"interrupt",...}
  • 接收:
    • chunk(增量片段):服务端 chunk_buffer 发送后即清零,前端直接 append_ai 拼接,为内容权威来源
    • complete(收尾信号):chunk 流式已完整渲染,complete 仅用于收尾;仅在完全丢帧(未收到任何 chunk)时用 full_text 补建消息,不再做全文替换对齐(避免服务端递归调用工具后 complete 只含最后一轮内容导致重复渲染)
    • line(过程行):🔧 工具调用提示、工具输出、💬 助手前缀,路由到任务侧栏;助手前缀/思考提示在聊天区已由 AI 消息框承载,不重复展示
    • reasoning(思考流):仅服务端 debug_level≥3 发送,记入任务侧栏
    • dialogue_ended / dialogue_reset / task_added / task_process / task_status / task_error / interrupt_ack / error / info / command_result / connection_ack / notification / pong