控制论反馈环

等待首轮数据
用户目标
等待首轮循环数据...
任务执行架构
graph TD subgraph G_dec[🧠 决策层] grok["葛老师 · Grok"] -->|写入决策| h2h[handoff-to-hermes.md] end subgraph G_exe[🤖 执行层] h2g[handoff-to-grok.md] -->|回执状态| grok hermes["Hermes Agent"] -->|读取决策| h2h hermes -->|执行 脚本/API/Git| bash[Shell / Python 脚本] bash -->|写入回执| h2g feishu[飞书] -->|用户消息| hermes end subgraph G_cron[⏱ 调度层 · 每20分钟] cron["Cron Job
check-and-sync.py"] -->|SHA256 检测变化| h2h cron -->|SHA256 检测变化| h2g cron -->|组装 JSON| jdata[feedback-loop-panel.json] jdata -->|scp 推送| k11_srv[(K11 服务器
nginx 静态文件)] end subgraph G_dash[📊 可视化层] dash["Dashboard
feedback-loop.html"] -->|fetch /state.json| k11_srv dash -->|30秒自动轮询| k11_srv dash -->|用户干预指令| clip[剪贴板] clip -->|粘贴到飞书| feishu end style grok fill:#1a3a5c,stroke:#4fc3f7,color:#4fc3f7 style hermes fill:#3a3a1c,stroke:#ffd54f,color:#ffd54f style cron fill:#1a3a2a,stroke:#81c784,color:#81c784 style k11_srv fill:#2a1a3a,stroke:#ce93d8,color:#ce93d8 style dash fill:#3a1a1a,stroke:#e57373,color:#e57373 style h2h fill:#0d0d1a,stroke:#4fc3f7,color:#4fc3f7 style h2g fill:#0d0d1a,stroke:#ffd54f,color:#ffd54f style jdata fill:#0d0d1a,stroke:#81c784,color:#81c784 style bash fill:#0d0d1a,stroke:#aaa,color:#aaa style clip fill:#0d0d1a,stroke:#e57373,color:#e57373 style feishu fill:#0d0d1a,stroke:#6fc,color:#6fc
当前循环
graph LR DEC[等待葛老师决策]:::wait --> EXE[等待执行]:::wait --> FB[等待反馈]:::wait FB -.-> DEC classDef wait fill:#1a1a2e,stroke:#444,color:#666 classDef active fill:#2a3a1e,stroke:#4fc3f7,color:#4fc3f7 classDef done fill:#1a3a2a,stroke:#81c784,color:#81c784 classDef block fill:#3a1a1a,stroke:#ff5252,color:#ff5252
循环历史(30天)
时间决策执行反馈
暂无数据
最近 handoff 原文
[葛老师] 等待首轮数据...\n [Hermes] 等待首轮数据...