manuscript-check

manuscript-check is a skill for Claude Code, Codex from MarkD1Zzz/workflow-tracker. It costs 124 tokens per session (2,243 once invoked), scanned A, original, MIT.

A Chinese-language skill for checking the completeness and consistency of academic manuscripts. It traces claims to original papers, code, and experiment records, then checks related text, tables, scripts, figures, and stored memories.

In plain words
What is it for?
Use it when verifying experiments, correcting manuscript data or terminology, changing tables, synchronising figures, or checking that the narrative remains consistent across files.
Why use it?
It helps catch unsupported data, incorrect attribution, stale numbers, inconsistent names, and mismatches between a paper and its plots or experiment code.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/markd1zzz/workflow-tracker/manuscript-check
Any agent
npx skills add MarkD1Zzz/workflow-tracker --skill manuscript-check
Clone the repo
git clone --depth 1 https://github.com/MarkD1Zzz/workflow-tracker

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for manuscript-check

README.md
[![agentmods](https://agentmods.dev/badge/skills/markd1zzz/workflow-tracker/manuscript-check.svg)](https://agentmods.dev/skills/markd1zzz/workflow-tracker/manuscript-check)
Your own site
<a href="https://agentmods.dev/skills/markd1zzz/workflow-tracker/manuscript-check"><img src="https://agentmods.dev/badge/skills/markd1zzz/workflow-tracker/manuscript-check.svg" alt="Measured on agentmods" height="20"></a>
Per session 124 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,243 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00124 $0.02243
Opus 5 $0.00062 $0.01122
Sonnet 5 $0.00025 $0.00449
Haiku 4.5 $0.00012 $0.00224

Measured 5d ago against content hash 10fa429cf4b7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

manuscript-check scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

manuscript-check/SKILL.md · 146 lines

How it starts

The opening of the file, as written. The whole thing — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Manuscript Integrity Checker

学术论文手稿的六步完整性校验闭环。当用户质疑手稿中的数据、命名、归属或叙事一致性时触发。

触发

检测到以下信号即开始执行:

  • 用户说"核查"/"核实"/"这个数据跑过吗"/"这个实验存在吗"
  • 用户指出归属错误("XXX 是我的工作,不是引用 YYY")
  • 用户指出命名错误("XXX 从未存在"/"这是旧稿遗留")
  • 用户指出数据不存在("没跑过"/"编的"/"这个消融没做")
  • 用户要求同步配图("图需要改吗"/"涉及图了吗")
  • 用户要求叙事统一("调整吧"/"不一致")

六步闭环

Step 1: 溯源核查

读取原始来源(原始论文/实验记录/代码),与当前手稿逐项对比。

操作: Read原始论文 → Grep关键术语 → 列出所有差异点
输出: 差异清单(位置、旧值、来源)

关键原则: 原始论文/代码是唯一真相来源。手稿必须与原始来源一致。

Step 2: 影响面分析

对每个差异点,搜索手稿中所有出现位置,评估修改波及范围。

操作: Grep -i 目标术语 → 列出所有行号 → 标注每处是否需要改、怎么改
覆盖范围: Abstract、§正文、Table、Figure caption、脚本文件(*.py)、实验日志(experiment_log.md)

特别注意脚本文件中的陈旧数值: 当表格数据被修改后,配图脚本中硬编码的 F1/Accuracy 值极易被遗漏。Step 2 必须对脚本中所有硬编码数值执行 grep,与当前表格逐行比对。

Step 3: 批量修改

按优先级一次性执行所有修改——先改 tex 正文,再改表格,最后改配图脚本。

表格修改注意事项:

  • 删除行后记得更新 \multirow{N} 中的 N
  • 删除缩写后更新 Table caption 的 Abbreviations

配图脚本注意事项:

  • 脚本中的数据必须与手稿表格完全一致
  • 修改后必须重新运行脚本生成图片

Step 4: 残留检查

修改完成后,grep 验证旧术语在全文(含脚本)中零命中。

操作: Grep -i 所有旧术语 → 确认零命中或仅保留位置正确
例外: 在原始论文中合理的术语可保留(如作为架构描述而非消融变量)

Step 5: 叙事一致性验证 + 脚本数据交叉校验

检查修改后的段落之间是否存在逻辑矛盾,同时交叉校验配图脚本中的硬编码数据是否与手稿表格一致

5a. 常见矛盾类型:

类型 示例 检查方法
数字矛盾 Table 说5行但正文说3个变体 手动逐行计数
术语矛盾 §4.1 说"普遍冗余"但 Conclusion 说"backbone-dependent" Grep 关键术语全文中所有出现
证据矛盾 消融数据不支持某结论但 benchmark 数据支持 区分语境(见下方"证据矛盾处理")
标签矛盾 Fig caption 描述与图中实际条形数不匹配 对照配图脚本中的 variants 列表
脚本数据陈旧 表格已改但 fig1_scatter() 中仍是旧 F1 值 (e.g. 0.9708 vs 应 0.9559) 逐函数 grep 硬编码数值,与表格逐行比对
对比不公平 fig_s1_sota 将 CE Loss 基线 (ConvNeXt=0.9904) 与 CB Focal Loss 本方法 (0.9727) 放在同一张图 检查每个 subplot 中数据点的 loss function/seed 来源

5b. 脚本数据交叉校验(强制执行):

1. Read 配图脚本中每个 def fig*() 函数
2. 提取函数内所有硬编码的 F1/Accuracy/参数值
3. 与手稿 Table 1/Table 2 逐值比对
4. 标记所有不一致 → 记录函数名、行号、旧值、应有值

5c. 证据矛盾处理: 当消融数据(Table 2)的结论与 benchmark 数据(Table 1)的结论表面冲突时,区分移除注意力和添加注意力的不同语境:

  • "移除 Transformer 注意力 → 零效应" = 注意力冗余(当已有强 backbone 特征时)
  • "向弱 backbone 添加 ViT → 正提升" = 注意力可替代弱 backbone 的不足
  • 两句均为真,但取决于不同基线条件 → 必须显式标注上下文

Read the full file on GitHub · 146 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago First seen · 146 lines · 124 tokens per session scan A 10fa429cf4b7

Subscribe to this mod's changes

manuscript-check is a skill published in the GitHub repository MarkD1Zzz/workflow-tracker (10 stars, last pushed 2mo ago), licensed MIT. It adds 124 tokens to every session and 2,243 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other skills, from other repositories

paper-narrative

Judge and reshape the STORY a paper's figures tell. Input is the work itself — manuscript (or abstract) + figure deck — no hand-written brief. paperbriefprompt(abstract, captions) hands you the prompt to write the brief yourself (pitch/vision/per-figure-claims); then you play a handling editor over the full deck and…

UnicomAI/wanwu · 155 tokens

tao-run-deft-cr-its-mining

Run the mining-based DEFT improvement workflow for ITS Cosmos-Reason binary video questions, focused on the non-reasoning classification/evaluation path. Use when the user asks for a DEFT CR ITS mining workflow, traffic-camera Cosmos Reason improvement loop, collision-identification workflow with data mining, or…

NVIDIA-TAO/tao-skill-bank · 80 tokens

tao-run-deft-object-detection

Run the full DEFT smart-data-augmentation loop for NVIDIA TAO Grounding DINO object detection: zero-shot baseline inference, KPI analysis, per-class gap analysis, SigLIP embedding of weak images, unique-neighbor mining against a source pool, ODVG dataset staging, and retraining — repeated for a fixed number of…

NVIDIA-TAO/tao-skill-bank · 196 tokens

sciverse-paper-search

Use this skill for scientific literature search, evidence retrieval, paper metadata screening, and cited research synthesis with Sciverse. This LazyLLM-adapted version supports SciverseSearch search, metasearch, metacatalog, and getcontent only; it does not assume full Sciverse MCP resource or attachment APIs are…

LazyAGI/LazyMind · 68 tokens

star-code-reviewer

Review code quality, docstrings, and whether a plan's code matches what it promised. No argument reviews all of ${CODENAME}/; a PLANNAME (slug / prefix / filename) reviews that plan's files plus conformance (§3 tasks, §4 deliverables, §5 done-criterion); a path reviews that path; diff or a git range reviews changed…

wanghao9610/STAR · 198 tokens

star-expt-digest

Summarize experiment progress since last time — the periodic digest a progress report is written from, in date order. No argument resumes from the last digest; a PLANNAME covers that node's whole family — ancestors for context, descendants for evidence; d or a date covers a window; all covers everything. Collects each…

wanghao9610/STAR · 205 tokens