agentic-sop-to-work

Turn human SOPs into deterministic, gated, human-approved agentic workflows — no fabrication, no "mega-agent" rot. 把人工 SOP 變成「確定性引擎 ＋ 誠實硬閘門 ＋ 人核准」的 agentic workflow——不臆造、不退化成 mega agent。

/plugin marketplace add s0912758806p/agentic-sop-to-work
/plugin install agentic-sop-kit@agentic-sop-to-work
/reload-plugins

🌐 English ・ 繁體中文

🗺️ At a glance · 運作一覽

Human SOP → single-tool skills → orchestrated flow (gated · branch · map) → DRAFT → human approval; a Stop-hook regression gate guards every change.Human SOP → 單一工具 skill → 編排流程（閘門・分支・map）→ DRAFT → 人核准；Stop-hook 回歸閘門守住每次變更。

flowchart LR
    SOP["📄 Human SOP"]
    SK["🔧 single-tool skills<br/>one tool · I/O contract"]
    RUN["⚙️ run.py<br/>steps · branch · map · --plan"]
    G{{"🚦 per-step gate<br/>cmd · schema · trace · recompute"}}
    DR["📝 DRAFT"]
    OK(["✋ human approval"])
    STOPN["✗ stop & fix"]
    REG[["🔁 Stop-hook regression<br/>change → tests → block & fix"]]

    SOP -- decompose --> SK
    SK -- orchestrate --> RUN
    RUN --> G
    G -- pass --> DR
    DR --> OK
    G -- fail --> STOPN
    REG -. guards every change .-> SK

    style OK fill:#E8F5E9,stroke:#2E7D32,color:#1B5E20
    style G fill:#FFF8E1,stroke:#F9A825
    style REG fill:#FFEBEE,stroke:#C62828,color:#B71C1C

English

What. agentic-sop-kit turns a process you do by hand (a "Human SOP") into an agentic workflow an LLM can run safely and repeatably. A methodology + portable toolkit — not a chatbot. Built for regulated / high-stakes / must-be-correct work.

Why it's safe — it blocks the predictable LLM failures:

• Fabrication → facts come only from inputs; gaps marked 【待補】, never invented.
• Fake autonomy → deterministic work in code; hard gates are hermetic & LLM-free (self-eval only advisory, capped).
• Unaccountable output → every output is a DRAFT; controlled / high-risk calls stay human-owned.
• Mega-agent rot → an audit skill + a Stop-hook regression gate that re-verifies on every change.

What you get

Engine (run.py) — deterministic, code-decided, additive:

• Per-step gates (block on fail, zero LLM): cmd_gate · schema_gate · trace_gate (no fabrication) · recompute_gate.
• cmd steps (allowlisted; mutations need --allow-mutations) · --plan dry-run (lists every step + validates branch gotos).
• branch (forward-only) · map_over (per-item) — control flow in code, never the model.
• Examples — dependency-free FE / BE / DB / AI flows in kit/workflow/examples/.
• Capped auto fix-loop — on a gate failure, /sop-flow auto-fixes & re-runs (run.py --max-fix-retries; code-enforced per run-id, capped by SOPKIT_MAX_FIX_RETRIES, default 3 — the same knob as the Stop-hook regression loop); exhausted → stop for a human; never patches output to pass.

Iron rules — facts only from inputs (【待補】 for gaps) · deterministic work in code, gates hermetic & LLM-free · DRAFT + human approval · gates check truth, not keywords. Real enforcement is the Stop-hook, not prose.

Install — in Claude Code (incl. Claude Desktop's Code tab):

/plugin marketplace add s0912758806p/agentic-sop-to-work
/plugin install agentic-sop-kit@agentic-sop-to-work
/reload-plugins      # or restart the session

Requires Python 3.8+ as python3 (macOS/Linux out of the box; on Windows ensure python3 resolves). Verify: /help lists /agentic-sop-kit:sop-flow.

Build your own flow — adopt the kit: python3 <plugin>/kit/bootstrap.py --project <dir> → scaffold a step: new_skill.py --name <x> (fill DEPS / run()) → wire workflow/flow.json → make it conversation-triggerable: export_claude_skill.py --skill <x>. Details: kit/SOP.md.

Layout

agentic-sop-to-work/
├── .claude-plugin/marketplace.json
└── plugins/agentic-sop-kit/
    ├── .claude-plugin/plugin.json · skills/ · commands/ · hooks/
    └── kit/   # portable kit: lib/ · workflow/ · tests/ · templates/ · examples/ · SOP.md

License — MIT (LICENSE · NOTICE); keep the copyright/license notice in copies, including single files. Updating: git push → users run /plugin marketplace update agentic-sop-to-work; bump plugin.json version on each meaningful change.

繁體中文

是什麼。 agentic-sop-kit 把你手動做的流程（Human SOP）工程化成 LLM 能安全、可重複執行的 agentic workflow。是一套方法論 ＋ 可攜工具包，不是聊天機器人。為受監管／高風險／不能錯的工作而生。

為何可信 — 逐一封堵 LLM 的可預期失敗：

• 臆造 → 事實只來自輸入；缺的標 【待補】，絕不杜撰。
• 假自主 → 確定性的事用程式；硬閘門 hermetic、零 LLM（自評只能 advisory 且封頂）。
• 無人負責的產出 → 產出一律 DRAFT；受控／高風險判定永遠由人擁有。
• mega-agent 退化 → 稽核 skill ＋ Stop-hook 回歸閘門，每次變更重新驗證。

你會得到

引擎（run.py） — 確定性、由程式決定、附加式：

• 每步硬閘門（fail 即停、零 LLM）：cmd_gate · schema_gate · trace_gate（防臆造）· recompute_gate。
• cmd 步驟（白名單；會改動環境的需 --allow-mutations）· --plan 乾跑（列出每步＋靜態驗證 branch goto）。
• branch（forward-only）· map_over（逐項）— 控制流由程式決定，不交給模型。
• 範例 — 免依賴的 FE／BE／DB／AI 流程：kit/workflow/examples/。
• 封頂自動修復 fix-loop——閘門失敗時 /sop-flow 自動修復並重跑（run.py --max-fix-retries，依 run-id 程式強制上限；與 Stop-hook 回歸共用 SOPKIT_MAX_FIX_RETRIES、預設 3）；用盡才停下交人；永不為過關竄改輸出。

鐵則 — 事實只來自輸入（缺標 【待補】）· 確定性用程式、閘門 hermetic 零 LLM · DRAFT ＋ 人核准 · 閘門查真相不查關鍵字。真正的強制力在 Stop-hook，不在散文。

安裝 — 在 Claude Code（含 Claude Desktop 的 Code 分頁）：

/plugin marketplace add s0912758806p/agentic-sop-to-work
/plugin install agentic-sop-kit@agentic-sop-to-work
/reload-plugins      # 或重開 session

需 Python 3.8+ 且能以 python3 呼叫（macOS／Linux 開箱即用；Windows 請確保 python3 可解析）。驗證：/help 應看到 /agentic-sop-kit:sop-flow。

建你自己的流程 — 導入 kit：python3 <plugin>/kit/bootstrap.py --project <dir> → scaffold 一步：new_skill.py --name <x>（填 DEPS / run()）→ 在 workflow/flow.json 接線 → 讓它對話可觸發：export_claude_skill.py --skill <x>。詳見 kit/SOP.md。

結構

agentic-sop-to-work/
├── .claude-plugin/marketplace.json
└── plugins/agentic-sop-kit/
    ├── .claude-plugin/plugin.json · skills/ · commands/ · hooks/
    └── kit/   # 可攜 kit：lib/ · workflow/ · tests/ · templates/ · examples/ · SOP.md

授權 — MIT（LICENSE · NOTICE）；副本需保留版權／授權聲明，含單一檔案。更新：git push → 使用者跑 /plugin marketplace update agentic-sop-to-work；每次有意義變更調高 plugin.json 的 version。