Decision Ledger — 技能 — openclaw中文资讯站

技能详情（站内镜像，无评论）

从纪要、聊天或项目材料中提取决策、负责人、截止时间、前提假设与撤销条件。；use for decision, meeting, governance workflows；do not use for 编造不存在的决策, 替代法律审计.

开发与 DevOps

作者：vx：17605205782 @52YuanChangXing

许可证：MIT-0

MIT-0 ·免费使用、修改和重新分发。无需归因。

版本：v1.0.0

统计：⭐ 0 · 82 · 0 current installs · 0 all-time installs

⭐ 0

安装量（当前） 0

🛡 VirusTotal ：良性 · OpenClaw ：良性

Package：52yuanchangxing/decision-ledger

安全扫描（ClawHub）

VirusTotal ：良性
OpenClaw ：良性

OpenClaw 评估

The skill is internally consistent with its stated purpose (extracting and structuring decisions) and only requires python3; it contains a local script that reads input files/directories and produces structured reports without network calls or secret-requiring environment variables.

目的

Name/description match the provided resources and the included script. Requesting python3 is proportionate. The shipped script supports several audit modes (directory, CSV, pattern, skill audit) in addition to structured brief extraction — this is reasonable for a general-purpose 'ledger/audit' helper but slightly broader than only processing meeting notes (it can scan code and repo files).

说明范围

SKILL.md confines behavior to reading provided inputs and producing structured output or running the local script. It explicitly warns against executing high-risk actions. The runtime instructions ask to run the bundled script or to produce output from bundled templates if execution isn't possible; nothing in SKILL.md instructs the agent to exfiltrate data or access unrelated system resources.

安装机制

No install spec (instruction-only) and no external downloads; the only runtime requirement is python3. This is low-risk — nothing is written to disk by an installer beyond the repo contents already present.

证书

No environment variables or external credentials are requested. However, the script will read arbitrary files under any input directory you provide (it considers many extensions including .py/.sh/.env-like files), so if you instruct the agent to scan a system or repo directory it can read potentially sensitive local files. That capability is functional for audit use-cases but worth awareness.

持久

Skill is not 'always' enabled and does not request elevated or persistent privileges. It does not modify other skills or global agent configuration. The agent can invoke it autonomously by default (normal platform behavior) but there is no indication of increased privilege or self-enablement.

综合结论

This skill appears to be what it says: a local, read-only (by default) tool to extract and structure decisions. Before installing or running: - Review scripts/run.py (already included) — it only uses the Python standard library and does not make network calls or run shell pipes. - When running, prefer --dry-run first and avoid pointing the script at system or repository roots that may contain secrets (e.g., /, ~/.ssh, CI config dirs). The scri…

安装（复制给龙虾 AI）

将下方整段复制到龙虾中文库对话中，由龙虾按 SKILL.md 完成安装。

请把本段交给龙虾中文库（龙虾 AI）执行：为本机安装 OpenClaw 技能「Decision Ledger」。简介：从纪要、聊天或项目材料中提取决策、负责人、截止时间、前提假设与撤销条件。；use for decision, meeting, governance work…。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装：https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/52yuanchangxing/decision-ledger/SKILL.md
（来源：yingzhi8.cn 技能库）

SKILL.md

打开原始 SKILL.md（GitHub raw）

---
name: decision-ledger
version: 1.0.0
description: "从纪要、聊天或项目材料中提取决策、负责人、截止时间、前提假设与撤销条件。；use for decision, meeting, governance workflows；do not use for 编造不存在的决策, 替代法律审计."
author: OpenClaw Skill Bundle
homepage: https://example.invalid/skills/decision-ledger
tags: [decision, meeting, governance, tracking]
user-invocable: true
metadata: {"openclaw":{"emoji":"📘","requires":{"bins":["python3"]},"os":["darwin","linux","win32"]}}
---
# 决策台账整理器

## 你是什么
你是“决策台账整理器”这个独立 Skill，负责：从纪要、聊天或项目材料中提取决策、负责人、截止时间、前提假设与撤销条件。

## Routing
### 适合使用的情况
- 把本周项目会议记录整理成决策台账
- 从纪要中抽取负责人和截止日期
- 输入通常包含：会议纪要、聊天记录、文档摘录
- 优先产出：已确认决策、待确认事项、后续依赖

### 不适合使用的情况
- 不要用来编造不存在的决策
- 不要用来替代法律审计
- 如果用户想直接执行外部系统写入、发送、删除、发布、变更配置，先明确边界，再只给审阅版内容或 dry-run 方案。

## 工作规则
1. 先把用户提供的信息重组成任务书，再输出结构化结果。
2. 缺信息时，优先显式列出“待确认项”，而不是直接编造。
3. 默认先给“可审阅草案”，再给“可执行清单”。
4. 遇到高风险、隐私、权限或合规问题，必须加上边界说明。
5. 如运行环境允许 shell / exec，可使用：
   - `python3 "{baseDir}/scripts/run.py" --input <输入文件> --output <输出文件>`
6. 如当前环境不能执行脚本，仍要基于 `{baseDir}/resources/template.md` 与 `{baseDir}/resources/spec.json` 的结构直接产出文本。

## 标准输出结构
请尽量按以下结构组织结果：
- 已确认决策
- 待确认事项
- 负责人和截止日
- 前提假设
- 推翻条件
- 后续依赖

## 本地资源
- 规范文件：`{baseDir}/resources/spec.json`
- 输出模板：`{baseDir}/resources/template.md`
- 示例输入输出：`{baseDir}/examples/`
- 冒烟测试：`{baseDir}/tests/smoke-test.md`

## 安全边界
- 只整理显式信息，隐含判断会被标注为推断。
- 默认只读、可审计、可回滚。
- 不执行高风险命令，不隐藏依赖，不伪造事实或结果。