Initial: multi-agent XMPP communication system with dashboard

- Platform-based architecture (Windows/Linux/Mac) - Agent instance registry (agents.yaml) - Management dashboard with cross-platform monitoring - xmpp_bot with HTTP bridge + health endpoints - wechat_agent with WeChat-Hermes bridging - Platform services: ProcessGuardian, HealthProbe, APIRouter, ChannelBridge - Deployment: systemd (Linux) + PowerShell (Windows) - Monitoring: SSH+ejabberdctl for cross-platform presence
2026-06-12 21:49:05 +08:00
commit 1b2b935832
76 changed files with 15943 additions and 0 deletions
@@ -0,0 +1,82 @@
+# AgentsMeeting — Session Handoff
+> 创建时间: 2026-06-12 04:00
+> 源会话: hermes session (ses_1d95d15c4ffehQaZ6hrbIbak5k)
+
+## 已完成
+
+### 系统修复
+- `part_` 前缀 bug (chat_bridge.py:335): `part_` → `prt_`
+- 工具调用 XML 泄漏: final force 用干净上下文 + `_strip_toolcall_xml()`
+- 静默模式误杀: `_SHUTUP_PATTERNS` 缩紧到仅 "闭嘴"/"住口"/"shut up"
+- self-message 回声静默: on_group_message 中 self-msg 检查提前到 shutup 之前
+- MAX_TOOL_LOOPS: 15 → 30，超限后走 clean final force
+- watchdog 进程唯一: `start_bot()` 先 `kill_all_bots()` 再启动
+- 健康检查: 检测 "alive 但无消息接收" 状态，自动重启
+- 日志轮转: watchdog 每 15 分钟检查 >5MB 自动轮转
+
+### 架构设计
+- docs/ARCHITECTURE.md (17.5 KB): 系统总览/拓扑/数据流/模块/监控
+- docs/AUDIT.md (4.2 KB): 稳定性审计报告
+- docs/DEPLOY.md (2.8 KB): 部署指南 (Windows + Linux)
+- docs/OPS.md (3.2 KB): 运维手册/常见故障/日志轮转
+- docs/QUICKSTART.md: 快速操作手册
+- README.md (4.4 KB): 项目概览
+
+### 部署脚本
+- deploy/windows/{start,stop,check,rollback}.ps1
+- deploy/linux/{install.sh,hermes-gateway@.service,xmpp-bot@.service}
+
+### 源码
+- src/shared/config.py: 集中配置管理 (env var + YAML)
+- src/shared/bot_base.py: Bot 基类 (BotConfig/BaseBot/extract_response)
+- src/ops/watch_group.py: 消息监控脚本
+- src/channels/qq/bridge.py: QQ 通道骨架
+- config/.env.example: 环境变量模板
+
+### HTTP 桥 (:5802)
+- POST /send: 发送群消息
+- GET /messages: 读取最近 200 条消息 (支持 ?from=mohe 过滤)
+- 集成到 xmpp_bot.py，bot 重启自动恢复
+
+### 测试
+- tests/test_core.py: 单元测试
+- tests/verify_deploy.py: 部署验证 (12 项全部通过)
+- HTTP bridge 验证: send+receive 均正常
+
+### 运行状态
+- xmpp_bot: RUNNING (watchdog 守护)
+- wechat_agent: RUNNING (uptime 35h+)
+- watchdog: RUNNING (30s 检查 + 日志轮转)
+- api_proxy: RUNNING
+- health_check: 5 分钟定时任务
+
+## 待完成
+
+### 需 mohe 在 Linux 侧
+- 部署 systemd services (hermes-gateway@.service, xmpp-bot@.service)
+- 配置 provider 环境变量 (替换 configs/*/config.yaml 中的占位符)
+- 排查 ejabberd MUC conference.yoin.fun 响应问题
+- state.db 829MB 归档清理
+
+### 需 mohe 配合
+- 架构文档最终审核
+- 部署验证
+- hermes-gateway@.service 使用实际 systemd 模板
+
+## 关键文件
+- projects/wechat-hermes-gateway/scripts/xmpp_bot.py (HTTP bridge)
+- projects/wechat-hermes-gateway/scripts/chat_bridge.py (final force 修复)
+- projects/wechat-hermes-gateway/scripts/xmpp_watchdog.py (log rotation)
+- projects/wechat-hermes-gateway/scripts/health_check_xxm.py
+- projects/AgentsMeeting/src/shared/config.py
+- projects/AgentsMeeting/src/shared/bot_base.py
+- projects/AgentsMeeting/deploy/windows/start.ps1
+- projects/AgentsMeeting/deploy/windows/check.ps1
+- projects/AgentsMeeting/docs/ARCHITECTURE.md
+- projects/AgentsMeeting/tests/verify_deploy.py
+
+## 注意事项
+- HTTP bridge (:5802) 绑定 127.0.0.1，不对外暴露
+- bot 重启后消息 buffer 会清空，新消息到达后自动重建
+- 健康检查 lastMsg=999min 是 MAM 恢复期间正常现象
+- MUC join 一直超时，bot 通过 MAM 订阅工作