Paper Iteration Pack logo

Paper Iteration Pack

Community
chengyike20110519-create
paper-iteration-pack

Iterate an existing paper or report from concrete review feedback: audit claims and source data, redesign figures, tighten layout, synchronize numbers, render, and verify the final document in place. Use for paper/figure revision packages, chart upgrades, and submission polish. Do not use for drafting a paper from scratch or for a one-line wording change with no evidence or layout impact.

Overview

Publisherchengyike20110519-create
RepositoryMathModelRun
Skill namepaper-iteration-pack
Stars
113
Forks
3
Bundled files
3
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 3 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by chengyike20110519-create on GitHub. Read the source before you install it.

Installation

Install the Paper Iteration Pack AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/chengyike20110519-create/MathModelRun/tree/main/skills/paper-iteration-pack
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/chengyike20110519-create/MathModelRun.git /tmp/MathModelRun
mkdir -p .claude/skills
cp -r /tmp/MathModelRun/skills/paper-iteration-pack .claude/skills/paper-iteration-pack
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Paper Iteration Pack in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Paper Iteration Pack on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Paper Iteration Pack is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Paper Iteration Pack

Use this skill when a paper, report, thesis section, competition paper, or slide-like document already exists and the user provides review suggestions, figure feedback, or a request to make the delivery clearer, tighter, brighter, or more submission-ready.

The unit of work is an iteration package, not a collection of isolated edits. One package normally contains:

  1. a source-of-truth ledger;
  2. the revised figure or layout source;
  3. synchronized text, captions, tables, and references;
  4. a rebuilt output artifact;
  5. visual and numeric verification;
  6. a short change and evidence report.

Non-negotiable rules

  • Read the existing document, current figure, generator, caption, and nearby claim before changing anything. Do not redesign from a screenshot alone.
  • Trace every changed number to a frozen result, run output, raw data, or an explicitly labeled user-provided value. Never silently reconcile a conflict by copying a number from the old figure.
  • Keep a stable rounding policy. A displayed percentage must agree with the same source value used in the abstract, body, table, figure, and conclusion.
  • Rebuild figures from their existing source when possible. Replace the target file in place; do not create v2, final-final, or parallel old versions.
  • Treat rendering as separate from generation. A saved PNG is not proof that the paper page is correct. Compile or export the document and inspect the actual page containing the revised material.
  • Record structural, runtime, and visual evidence separately. A file-existence check does not prove that a chart is readable or that a PDF page fits.
  • Preserve the author's language, terminology, citation style, and document class unless the review request explicitly asks for a change.
  • Stop and report a blocker when source data, generator code, or required rendering tools are missing. Do not invent data to make the visual look complete.

Iteration workflow

  1. Frame the requested change. Extract the review suggestions and turn them into observable acceptance criteria: what must change, where it must appear, and how it will be checked.
  2. Audit the current state. Locate the document source, target page or section, figure file, figure generator, source data, caption, and every repeated number or claim.
  3. Choose the revision scope. Decide whether this is a figure-only change, figure plus text, layout-only change, evidence/provenance repair, or a source-data problem that requires upstream regeneration.
  4. Choose the visual encoding. Match the chart to the comparison. Read references/visual-selection.md when the request mentions advanced charts, bar-chart upgrades, density, color, or 3D.
  5. Implement in the canonical source. Patch the existing plotting or layout code, regenerate the target artifact in place, and keep any style preset configurable rather than hard-coding one paper's content.
  6. Synchronize the document. Update the caption, body, table, abstract, labels, and evidence mapping whenever the visual or source value changed.
  7. Render and inspect. Rebuild the paper or report and inspect the actual revised page at normal reading size. Check clipping, overlap, font size, contrast, aspect ratio, aspect-ratio distortion, and print legibility.
  8. Report the package. State changed files, source values, visual changes, commands or tools used, verification level, and any unresolved limitation.

For a fuller stage-by-stage procedure, read references/iteration-workflow.md.

Figure decisions that are easy to get wrong

  • Use a grouped bar or dot plot when a small number of exact category values must be compared. Label values directly and start the scale at zero for bars.
  • Add a second panel only when it answers a distinct question. A useful pair is absolute values plus relative change, a slope chart, or a small-multiple breakdown.
  • Prefer dumbbell or slope charts for before/after comparisons. Prefer a lollipop or ordered dot plot when many categories make solid bars noisy.
  • Use a heatmap for a matrix, a response surface for two continuous inputs and one continuous response, and a spatial plot when the data have real geometry.
  • Do not use 3D merely to make a chart look advanced. A 3D bar chart normally hides exact values and adds occlusion. Use 3D only when the third dimension is a genuine ordered or continuous variable and rotation reveals structure that a 2D projection cannot communicate reliably.

MCM/MCM-style visual preset

Treat this as a preset, not a universal rule:

  • compact canvas with controlled margins;
  • bright but distinguishable colors with sufficient contrast;
  • light neutral panel background, subtle grid, no heavy border;
  • direct labels or direct annotations instead of a distant legend when space permits;
  • one result-oriented title per panel or figure;
  • full-width placement only when the extra detail is readable at paper scale.

Do not describe a figure as submission-ready until the rendered page has been inspected. If the renderer or page inspection is unavailable, say exactly which part remains unverified.

Completion report

Keep the final report short and evidence-oriented. Include:

  • Changed: document, figure, caption, data, and generator paths;
  • Data: source files or frozen values used, including units and rounding;
  • Visual decision: what was changed and why that encoding fits the claim;
  • Verification: generation command/result, render command/result, and what was visually inspected;
  • Remaining risk: any missing source, tool limitation, or unchecked claim.

Use the user's language for the report. In Chinese requests, use 中文为主、英文术语保留原文 unless told otherwise.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Paper Iteration Pack AI skill do?

Iterate an existing paper or report from concrete review feedback: audit claims and source data, redesign figures, tighten layout, synchronize numbers, render, and verify the final document in place. Use for paper/figure revision packages, chart upgrades, and submission polish. Do not use for drafting a paper from scratch or for a one-line wording change with no evidence or layout impact.

Why use Paper Iteration Pack on TypingMind?

Because you install it once and use it with any model. Paper Iteration Pack is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Paper Iteration Pack in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/chengyike20110519-create/MathModelRun/tree/main/skills/paper-iteration-pack. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Paper Iteration Pack?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Paper Iteration Pack?

As many as you like. As long as a model supports skills, you can use Paper Iteration Pack with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Paper Iteration Pack AI skill free?

Yes. It is published on GitHub by chengyike20110519-create under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇