Skip to content
Agent Orchestration
Skill

/accept

Use as a terminal node that emits accepted@1 only when an upstream verdict@1 says pass, and fails loudly on a rejected draft regardless of how routing reached it. Single tool. Reusable standalone.

BOOST
From plugin
agentic-sop-to-work
20911 skills4 commands
Install
$ npx -y skills add s0912758806p/agentic-sop-to-work --skill accept --agent claude-code

How it fires

How this skill gets triggered: by you, by Claude, or both.

  • Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
  • You can call itInvoke it directly when you want it.
  • Slash command/accept

Context preview

The summary Claude sees to decide when to auto-load this skill.

Use as a terminal node that emits accepted@1 only when an upstream verdict@1 says pass, and fails loudly on a rejected draft regardless of how routing reached it. Single tool. Reusable standalone.

SKILL.md

accept.SKILL.md
name: accept
description: Use as a terminal node that emits accepted@1 only when an upstream verdict@1 says pass, and fails loudly on a rejected draft regardless of how routing reached it. Single tool. Reusable standalone.

Skill — accept(終端節點,不信任路由)

圖示範流程的終端節點。刻意**不信任路由**:即使編排把一份 `reject` 的判定送到這裡 (例如有人手改了 `flow.json` 的 branch 條件),這一步仍拒絕通過。 閘門查真相,不查「我是被誰叫來的」。

綁定的單一工具

  • `python3`(標準庫)。**本 skill 只此一個工具。**

依賴

  • `python` >= 3.8

參數化(無硬編碼)

  • 輸入:`--in <verdict artifact.json>`;輸出:`--out <accepted artifact.json>`

介面

  • **輸入 artifact**:`verdict@1`(讀 `data.verdict`、`data.round`)。
  • **輸出 artifact**:`accepted@1` — `data{accepted, rounds, draft_status:"DRAFT"}`,透傳 `trace`。

fail-loud 邊界

`data.verdict != "pass"` → **exit 3**。產出一律 DRAFT,需人覆核。

執行

`python3 skills/accept/tool.py --in <$RUN/verdict.json> --out <$RUN/accepted.json>`

獨立重用

`skills/accept/` + `lib/kit.py` 可單獨抽出;只要上游給 `verdict@1` 即可運作。

Ships withagentic-sop-to-work

Loop Engineering — make an agent's loop produce verifiable progress, stay honest, and stay bounded, by deterministic mechanism rather than vibes.

Get the whole plugin

Other skills on agentic-sop-to-work.