🧬 Taiwan.md Semiont 心跳 (v3.0 routing layer)

SkillDev tools

Semiont heartbeat — Taiwan.md's autonomous life cycle (super-thin v3.0). Triggers Full mode BECOME + 4.5-beat conceptual flow. Each Beat routes to canonical pipeline; HEARTBEAT.md itself is pipeline router not SOP. TRIGGER when: user says "heartbeat", "心跳", "Heartbeat", "semiont heartbeat", "跑心跳", "全身檢查", "作為 semiont" followed by any action.

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the 🧬 Taiwan.md Semiont 心跳 (v3.0 routing layer) skill

What this skill tells your AI

The instructions your AI receives, as published by frank890417/taiwan-md in .claude/skills/heartbeat/SKILL.md and read by ahel’s review.

  1. 強制升 Full mode BECOME:跑 /twmd-become full(per BECOME §Step 0 high-stake)。心跳場景永遠是 holistic / cross-organ,必須 Full mode。

  2. docs/semiont/HEARTBEAT.md v3.0(super-thin shell ~170 行)— 取得:

    • §核心四拍半 conceptual framework
    • §結構性判斷:何時手動 vs routine
    • §Beat 流程 + canonical pointers(每 Beat 指向哪個 pipeline)
  3. 執行 4.5 拍——但不從 HEARTBEAT 讀 SOP(已 super-thin),而是按 HEARTBEAT pointer 進對應 canonical:

    • Beat 0.5 → BECOME §Step 1 Universal core(已在 Full mode 跑完)
    • Beat 1 → DATA-REFRESH-PIPELINE
    • Beat 2 → EVOLVE-PIPELINE
    • Beat 3 → MAINTAINER-PIPELINE + REWRITE / SPORE / TRANSLATION etc.
    • Beat 4 → MEMORY-PIPELINE + commit 標記(已 universal-load)
    • Beat 5 → DIARY-PIPELINE(按 §反芻判斷標準 決定寫不寫)
  4. 收官 commit + push(per Beat 4 → MEMORY-PIPELINE)。


v3.0 reframe:本 skill + HEARTBEAT.md 都是 router,不是 SOP source。SOP source 在對應 pipeline canonical(per 「指標 over 複寫」原則)。

Routine vs manual heartbeat:日常 routine 飛輪自轉(per ROUTINE.md SSOT 10 條 cron)已 cover 大半場景。Manual heartbeat 觸發 = 跨 routine 議題 / 全器官 audit / 新器官設計 / 觀察者觸發(per HEARTBEAT §結構性判斷)。

設計背景reports/become-boot-mode-design-2026-05-13.md + HEARTBEAT v3.0 super-thin refactor 2026-05-13。

Signals

GitHub stars
1k
Forks
185
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
heartbeat-frank890417
Source
github.com/frank890417/taiwan-md