技能详情（站内镜像，无评论）
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 · 38 · 0 current installs · 0 all-time installs
⭐ 0
安装量（当前） 0
🛡 VirusTotal ：良性 · OpenClaw ：良性
Package：52yuanchangxing/evidence-gap-mapper
安全扫描（ClawHub）
- VirusTotal ：良性
- OpenClaw ：良性
OpenClaw 评估
The skill's code, instructions, and requirements are coherent with its stated purpose (local evidence-gap analysis); it requires only python3, contains a local script that analyzes files you point it at, and does not attempt network exfiltration or request unrelated credentials.
目的
Name and description match the included assets: SKILL.md, resources/spec.json/template.md, and scripts/run.py implement local evidence-gap analysis, directory/csv/pattern audits, and structured brief generation. Declared requirement (python3) is proportional.
说明范围
SKILL.md correctly instructs running the local script or falling back to templates. The script reads the input path (file or directory) and samples many text file types (.md, .py, .csv, .sh, etc.) to produce reports; this is expected for a local audit tool. Caution: if you pass broad system paths (e.g., /, ~, or your repo root) the script will read and summarize those files and may surface snippets that look like secrets (it attempts partial r…
安装机制
No install spec is provided (instruction-only with an included helper script). That is low-risk: nothing is downloaded from external URLs and no packages are installed automatically. The only runtime requirement is a local python3 interpreter.
证书
No environment variables, credentials, or config paths are requested. The script operates on user-supplied input paths only, so requested privileges are minimal and appropriate.
持久
The skill is not set to always:true and does not request persistent system or agent-level changes. It only writes output when you specify an --output path (or the agent chooses to), and SKILL.md emphasizes read-only / audit-first behavior.
综合结论
This skill appears to do exactly what it claims: local, template-driven evidence-gap analysis using a Python helper script. Before running it, review the included scripts (scripts/run.py) yourself. Only pass the files or directories you intend it to analyze — avoid pointing it at entire system roots, home directories, or other places containing secrets. The script will read text files you point it to and may surface snippets (it masks long sec…
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「Evidence Gap Mapper」。简介：在报告、方案或演示稿中定位结论先行但证据不足的位置，并给出补证优先级。；use for evidence, gap-analysis, research wo…。
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/evidence-gap-mapper/SKILL.md
（来源：yingzhi8.cn 技能库）SKILL.md
---
name: evidence-gap-mapper
version: 1.0.0
description: "在报告、方案或演示稿中定位结论先行但证据不足的位置，并给出补证优先级。；use for evidence, gap-analysis, research workflows；do not use for 伪造数据支撑结论, 忽略高风险假设."
author: OpenClaw Skill Bundle
homepage: https://example.invalid/skills/evidence-gap-mapper
tags: [evidence, gap-analysis, research, quality]
user-invocable: true
metadata: {"openclaw":{"emoji":"🕳️","requires":{"bins":["python3"]},"os":["darwin","linux","win32"]}}
---
# 证据缺口绘图师
## 你是什么
你是“证据缺口绘图师”这个独立 Skill，负责：在报告、方案或演示稿中定位结论先行但证据不足的位置，并给出补证优先级。
## Routing
### 适合使用的情况
- 找出这份报告里证据不足的地方
- 给我一个补证优先级
- 输入通常包含：文稿、结论、现有证据
- 优先产出：主要结论、证据现状、下一步
### 不适合使用的情况
- 不要伪造数据支撑结论
- 不要忽略高风险假设
- 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：
- 主要结论
- 证据现状
- 缺口列表
- 补证优先级
- 可降级表述
- 下一步
## 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
- 适合作为质检器使用。
- 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。