thesis-deai-review SKILL.md skillA
thesis-deai-review from gxgeek-n/wenqu-mem is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.
mdr add gxgeek-n/wenqu-mem/thesis-deai-review@git:20260807.c0a57aaPins git:20260807.c0a57aa in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 17 of 17 checks passed
- Kind
- SKILL.md skill
- Repository
- gxgeek-n/wenqu-mem
- Path
skills/thesis-deai-review/SKILL.md- Versions recorded
- 5
- Latest version
- git:20260807.c0a57aa, committed Aug 7, 2026
- First version recorded
- Jul 4, 2026
- Size
- 9,840 bytes
- Repository stars
- 4
- License
- a license file GitHub does not identify
审查中文学位论文（尤其 MEM/工程管理类实证改进论文）的模板化表达、证据薄弱点和"AI 味"风险，让它回到真实材料和专业硕士论文语体。 核心是以真实通过的同类论文为体裁基准，再让确定性扫描脚本与隔离 AI 判断并行，合并证据后定向修改。 当用户问"像不像 AI 写的""哪里太模板化""证据链哪里薄""脚本和AI并行审查""润色得像学生论文"时使用。 分工（表达风险三角）：正向语体写法（句式/引号/段落形态）→ thesis-style；AIGC 检测报告 （PaperYY/PaperPass/知网 报告解析+复核闭环）→ aigc-reduce-playbook；本 skill 专做"像不像 AI"的**判定** 与定向修改边界，不承诺平台结果。.
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# thesis-deai-review —— 论文模板化表达与 AI 味风险审查 把一篇中文学位论文中“过度模板化、证据链太薄、论证太顺滑”的位置找出来，回到真实材料和专业硕士论文语体上修。 **不是**做表面文字替换，也**不是**改成大白话，更不是承诺任何检测平台给出特定结果。 ## 一条最重要的原则（本方法论的核心） > **先找一篇同专业、同体裁、已通过答辩的真人论文当"基准"，用它校准"什么是 AI 痕迹"。** 很多"看起来像 AI"的特征，其实是**体裁规定动作**，真人论文一模一样——改掉反而更不规范、更可疑。 实测对照（MEM 质量改进类论文）发现，以下**全是体裁规范，不要动**： - 三个改进章高度平行（问题→原因→方案→实施→效果→本章小结） - "本章小结"复述式（"本章围绕……完成了……下一章将……"） - 摘要以"随着……"开头、六段式（背景→对象→方法→问题→结果→结论） - 程度副词"有效/显著/全面/系统"高频（真人论文往往用得**更多**） - 结论"研究结论 + 研究展望"二段式、理论/标准映射表、方案多标准优选表 ## 真正的 AI 指纹（这才是要改的） 针对**实证/数据驱动**的改进类论文，AI 味集中在"数字和论证太完美"，而非"空话连篇"： 1. **统计过度武装且全部完美** ★最强信号★：几乎每个指标都做 t 检验，且 **全部 p<0.001、全部 |Cohen's d|>1.0、95%CI 全部不含 0**；甚至对"系统埋点的 总体计数"（如某率、某分位延迟等）也套样本 t 检验。真人论文几乎不会这样。 2. **叙事与数据表自相矛盾**：正文写了波折（如大促性能回落、灰度超标、试点指标短暂下跌）， 但趋势表/效果表却一条平滑单调线——说明"叙事"和"数据"是两套流程分别生成、没对账。 3. **数字过于工整**：满意度全挤在某个窄区间（如 4.0–4.4）、占比恰好凑成 80%/100%、 权重恰好是 0.05 的整数倍、效应量随基线"完美单调对应"。 4. **诊断指标教科书式整齐**：回归里 VIF 全部刚好 <5、R² 是整数、Shapiro p 刚好不拒绝、 Cook's 远小于阈值——真实回归通常至少有一项处于临界、有变量因共线性掉出来。 5. **"直接给方案"无权衡**：方案设计直接给"方案一/二/三"，没有备选、没有取舍、没有优选过程。 6. **空泛万能句**：换个公司/行业也成立的过渡句、"为……提供借鉴"式安全表态（少量是体裁，扎堆才是问题）。 ## 去 AI 味的标准动作（保持正式学术书面语！） > ⚠️ **最大的坑**：去 AI 味 ≠ 改成口语。把"本章完成了……"改成"回头看，这章踩了几个坑" …
Read the whole file at its exact version.
How to install
mdr add gxgeek-n/wenqu-mem/thesis-deai-review@sha256:7225090934ada019The line at the top of this page pins the label git:20260807.c0a57aa, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_azm5b4i5u6q56hmy)
0 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 5 versions of this file, between Jul 4, 2026 and Aug 7, 2026, about one every 9 days.
| version | committed | commit | size | audit | |
|---|---|---|---|---|---|
| git:20260807.c0a57aa latest | 2026-08-07 | c0a57aa | 9,840 B | A | view · diff |
| git:20260726.6c7e134 | 2026-07-26 | 6c7e134 | 9,823 B | A | view · diff |
| git:20260704.1ab6c99 | 2026-07-04 | 1ab6c99 | 9,865 B | A | view · diff |
| git:20260704.eb9bb8b | 2026-07-04 | eb9bb8b | 9,900 B | A | view · diff |
| git:20260704.d00de62 | 2026-07-04 | d00de62 | 9,601 B | A | view |
Audit of the latest version
- pass: Frontmatter block present
- pass: Frontmatter declares a name
- pass: Frontmatter declares a description
- pass: Size between 200 bytes and 200 KB (9840 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
gxgeek-n/wenqu-mem · 4 stars · license a license file GitHub does not identify · pushed 2026-11-08 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_azm5b4i5u6q56hmy GET https://markdownregistry.com/api/v1/resolve?ref=gxgeek-n/wenqu-mem/thesis-deai-review GET https://markdownregistry.com/api/v1/blob/7225090934ada01945a57861c84c3f8a417f1146e24a18471387e21bbac1afe4
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.