Figure Table Audit logo

Figure Table Audit

Community
scdenney
figure-table-audit

Runs end-stage QA on a finished figure and table set. Checks callouts, numbering, and cross-references, whether numbers in the text match the exhibits, self-contained captions and statistical notes, accessibility and production quality, and links to the SI and replication package. Values that could only be read off an image are marked for author verification rather than estimated. Use when preparing a submission. Design guidance during drafting belongs to figures and tables.

Overview

Publisherscdenney
Repositoryopen-science-skills
Skill namefigure-table-audit
Stars
69
Forks
3
Bundled files
Instructions only
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.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by scdenney on GitHub. Read the source before you install it.

Installation

Install the Figure Table Audit 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/scdenney/open-science-skills/tree/main/plugin/skills/figure-table-audit
  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/scdenney/open-science-skills.git /tmp/open-science-skills
mkdir -p .claude/skills
cp -r /tmp/open-science-skills/plugin/skills/figure-table-audit .claude/skills/figure-table-audit
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Figure Table Audit 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 Figure Table Audit 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 Figure Table Audit 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.

Figure and Table Auditor

The end-stage auditor for a stable figure and table set, run before submission. Design and production guidance during drafting lives in figures and tables. When a claim depends on values plotted in an image, check it against the source data, or mark it for author verification; do not estimate values by eye.

Adapted for open-science social-science manuscripts from figure, table, and citation-compliance ideas in Cheng-I Wu's Academic Research Skills for Claude Code (CC BY-NC 4.0).

Instructions

1. Orient before auditing

Identify:

  • Manuscript source and compiled output, if any.
  • Figure directory, table files, appendix/SI files, and build command.
  • Target journal or style guide.
  • Whether figures are generated from code, manually edited, or exported from software.
  • Whether source data for figures/tables are available.

If only a PDF is available, state that cross-reference and value checks are lower confidence.

2. Inventory figures and tables

Build an inventory with:

  • Figure/table number or label.
  • File path or source location.
  • Caption/title.
  • First in-text callout.
  • Appendix/SI location if applicable.
  • Producing script or data source, if visible.

Check:

  • Every in-text callout resolves to an existing figure/table.
  • Every figure/table is called out in text before or near appearance.
  • Numbering is gap-free and not duplicated.
  • Main-text and SI labels do not collide.
  • LaTeX/Markdown references compile without unresolved labels.

3. Check text-to-evidence consistency

For each figure/table used to support a substantive claim:

  • Match the in-text claim to the exact row, column, panel, model, or plotted quantity, and name it in the finding.
  • Verify direction, magnitude, uncertainty, subgroup, and denominator.
  • Check whether text overstates non-significant or imprecise estimates.
  • Check whether figure/table notes disclose model specification, N, weights, fixed effects, clustering, and missing-data handling.
  • For experimental papers, check ITT vs per-protocol labeling, attrition-by-arm, baseline balance, and CONSORT/sample-flow consistency.

Read exact values from a plot only when the figure labels them. If source data are unavailable, write VISUAL READ ONLY - AUTHOR VERIFY.

4. Audit captions and notes

Captions and table notes should let a reader understand the evidence without hunting:

  • Figure caption states what is plotted, units, sample, uncertainty interval, and panel meanings.
  • Table title states the estimand or model family, not only "Results."
  • Table notes define dependent variable, treatment/condition coding, omitted categories, controls, fixed effects, clustering, weights, and significance markers.
  • All abbreviations are defined on first use.
  • All transformations, scales, and index directions are explicit.
  • Any data exclusions or subgroup restrictions are named.

For conjoint, list-experiment, topic-modeling, text-classification, and OCR studies, invoke or recommend the relevant sibling skill when table/figure interpretation depends on method-specific standards.

5. Check accessibility and production quality

Flag:

  • Color palettes not interpretable in grayscale or by color-blind readers.
  • Low-resolution or rasterized text in vector outputs.
  • Axis labels too small or missing units.
  • Legends that obscure data or use ambiguous labels.
  • Inconsistent decimal precision.
  • Overcrowded tables that should move to SI.
  • Missing alt text when the target venue requires it.
  • Non-reproducible manual edits that are not documented.

6. Audit SI and replication linkage

Check whether:

  • Main-text claims that point to SI land on the correct appendix/table/figure.
  • SI numbering and captions are internally consistent.
  • Each main figure/table can be traced to a script or documented manual step.
  • Replication README maps outputs to scripts and data files.
  • Figure/table source data are archived or clearly restricted.

Output

Produce a Figure and Table Audit Report:

# Figure and Table Audit Report

Scope:
Inputs checked:
Build/source status:
Summary: <N blocking, N recommended, N minor, N author-verification>

## Inventory
| ID | Path/location | Caption/title | First callout | Source/script |

## Blocking Issues
| Location | Figure/table | Issue | Evidence | Fix |

## Recommended Fixes
| Location | Figure/table | Issue | Fix |

## Minor / Production Issues
| Figure/table | Issue | Fix |

## Author Verification Needed
| Figure/table | Why verification is needed |

## Readiness Checklist
| Dimension | PASS/FAIL/PARTIAL/NA | Notes |

Severity:

  • Blocking: missing figure/table, wrong referenced value, denominator mismatch, unresolved label, table contradicts text, missing sample-flow evidence for experimental paper, or non-reproducible main result output.
  • Recommended: incomplete notes, unclear captions, missing units, missing source script, accessibility problem, imprecise uncertainty reporting.
  • Minor: style, spacing, decimal precision, typography, cosmetic consistency.

Frequently asked questions

What does the Figure Table Audit AI skill do?

Runs end-stage QA on a finished figure and table set. Checks callouts, numbering, and cross-references, whether numbers in the text match the exhibits, self-contained captions and statistical notes, accessibility and production quality, and links to the SI and replication package. Values that could only be read off an image are marked for author verification rather than estimated. Use when preparing a submission. Design guidance during drafting belongs to figures and tables.

Why use Figure Table Audit on TypingMind?

Because you install it once and use it with any model. Figure Table Audit 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 Figure Table Audit in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/scdenney/open-science-skills/tree/main/plugin/skills/figure-table-audit. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Figure Table Audit?

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 Figure Table Audit?

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

Is the Figure Table Audit AI skill free?

It is published on GitHub by scdenney. Check the repository for licensing terms. 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 👇