Works with the AI Agents you already use
Claude Code
Cursor
Codex CLI
Muse
Dots+21 moreClaude CodeCursorCodex CLIMuseDots
Manus
Grok
Grok Bot+18 agentsChart & Statistics Integrity Gate Skill
Adversarial audit tool for detecting misleading charts, statistical fallacies, and unsupported data claims.
$9.99
Chart & Statistics Integrity Gate Skill
Example session with this skill installed
Audit this claim: 'User engagement is up 50% this month!' based on these stats: Month A: 100 users, 10 posts. Month B: 150 users, 11 posts. The chart starts its Y-axis at 10 posts.
- Read your context and instructions
- Compiled the chart & statistics
The audit is complete. The presentation of these statistics is materially misleading due to denominator neglect and visual scaling issues.
Integrity Verdict: FAIL
Executive Findings
The claim of a 50% increase in 'engagement' relies on total user growth while ignoring that engagement per user has actually declined. The visual representation compounds this by truncating the Y-axis to exaggerate a marginal increase in total activity.
Issue Severity Table
| Issue | Severity | Type | Description |
|---|---|---|---|
| Denominator Neglect | Critical | Statistical | 50% user growth is conflated with 50% engagement growth. |
| Engagement Rate Decline | High | Inference | Posts per user dropped from 0.10 to 0.07 (a 30% decrease). |
| Truncated Y-Axis | Medium | Visual | Starting Y-axis at 10 makes the move from 10 to 11 look significant. |
Required Corrections
Redefine Metric: Change the headline to 'Total Users increased 50%' or report 'Engagement per User decreased 30%'.
2. Adjust Baseline: Set the Y-axis baseline to 0 to show the 10% increase in total posts in its true context.
3.
Contextualize Engagement: Include the denominator (total users) in the engagement calculation to avoid misleading the audience.
Next steps
- Recalculate engagement as a rate (posts/user) before publishing.
- Update the chart Y-axis to start at zero.
- Revise the headline to reflect user growth rather than engagement quality.
Connects securely to your tools. The creator never sees your data.
About this skill
Data visualizations and statistical summaries often hide critical flaws, whether through cherry-picked axes or confusing correlation with causation. This skill acts as an adversarial auditor to catch misleading charts and unsupported claims before they reach your stakeholders or the public.
What it does
- Visual Integrity Auditing identifies truncated axes, misleading scales, dual-axis distortions, and area exaggerations.
- Statistical Logic Checking flags mean vs. median misuse, percentage-point confusion, and Simpson's Paradox risks.
- Denominator Analysis investigates if rates are compared fairly and if the underlying population context is missing.
- Causality Verification separates descriptive facts from unsupported causal claims and identifies potential confounding variables.
- Severity Assessment categorizes issues from "Informational" to "Critical" based on their potential to materially mislead.
How it works
- Submit Material by providing a chart description, data table, or a summary of statistical claims.
- Audit Execution through visual, statistical, context, and inference review phases.
- Review Verdict including a PASS, REVISE, or FAIL status based on identified distortions.
- Implement Corrections using the provided list of specific recommendations for labels, scales, or wording.
Frameworks & tools
This skill is designed for data analysts, journalists, and researchers using tools like Excel, Tableau, PowerBI, D3.js, or Matplotlib who need a rigorous peer review of their output's integrity.
Why this beats prompting it yourself
Generic prompts often focus on making charts look "better" rather than identifying why they are "wrong." This skill uses a structured, adversarial checklist that specifically looks for survivorship bias, denominator neglect, and visual trickery that standard LLM instructions miss.
Use cases
- Internal Reporting to ensure quarterly KPIs are not presented with misleading baselines.
- Data Journalism to verify that a source's infographic isn't cherry-picking a specific timeframe.
- Academic Review to check if a paper's summary statistics match the reported sample size and methodology.
- Marketing Compliance to audit claim strength and ensure "percent increase" labels are statistically sound.
Known limitations
Does not perform raw calculation of new statistical tests if data is missing. Requires a text-based description or raw data of the visual elements to audit them effectively.
Cannot execute new statistical tests or generate missing raw data. Requires detailed descriptions of visual encodings if analyzing image-based charts without underlying data tables.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 1 month ago
Permissions
No special permissions declared or detected