agentsop-bio-fraud-forensics

Screen biomedical papers for image and statistical anomalies using forensic checklists.

287|16|Updated May 20, 2026
One-click install
npx skills add https://github.com/agentsope/SkillAlchemy --skill agentsop-bio-fraud-forensics
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agentsop-bio-fraud-forensics
Source: https://github.com/agentsope/SkillAlchemy/tree/main/skills/agentsop-bio-fraud-forensics
Command: npx skills add https://github.com/agentsope/SkillAlchemy --skill agentsop-bio-fraud-forensics

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It helps users screen biomedical and life-science papers for observable signs of image/data manipulation and statistical anomalies, while keeping outputs as neutral, evidence-based questions rather than accusations.

Core Features & Use Cases

  • Reproducible integrity screening: Produces a structured per-paper checklist focused on what can be seen and verified from the published artifact.
  • Image forensics workflow: Guides image-panel triage and verification using duplication categories (Bik Type I/II/III) and transformation-confirmed overlays (flip/rotate/compare).
  • Statistical consistency checks: Recomputes or validates reported statistics (e.g., GRIM/GRIMMER/statcheck-style logic) and flags mathematical impossibilities or inconsistencies.
  • Verification routing: Suggests where and how to check status and prior concerns (PubPeer, Retraction Watch/ORI, PubMed/Crossmark, scite) without treating status lookups as the core method.
  • Safety/defamation guardrails: Enforces graded language (Tier 1 observable anomalies, Tier 2 author-clarification questions) with a “benign-explanation gate” before escalation.

Quick Start

Ask: “Help me screen this biomedical paper (or this figure/blot/DOI) for observable image duplication, blot splicing, and/or statistical anomalies, and write your findings as reproducible Tier 1–Tier 2 questions with neutral wording.”

Frequently Asked Questions about agentsop-bio-fraud-forensics

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check a biomedical paper for image duplication and statistical anomalies?

To check a biomedical paper for image duplication and statistical anomalies, submit the target DOI, figure image, or .xlsx supplement to apply field-distilled forensic checklists that produce transformation-confirmed observations.

What is the best way to screen a Western blot panel for evidence of manipulation?

Screening a Western blot panel for manipulation involves guiding image-panel triage using Bik Type I/II/III duplication categories and applying transformation-confirmed overlays like flip, rotate, and compare to flag observable anomalies.

How do I validate reported statistics in a life-science paper supplement?

To validate reported statistics in a life-science supplement, the screening process recomputes numeric summaries using GRIM/GRIMMER/statcheck-style logic to flag mathematical impossibilities or statistical inconsistencies.

Can I verify suspected research integrity red flags without making defamatory accusations?

You can verify suspected research integrity red flags safely because the screening enforces graded, non-accusatory language with a benign-explanation gate, producing Tier 1 observable anomalies and Tier 2 author-clarification questions.

Where should I look to verify if a biomedical paper has been retracted or flagged on PubPeer?

To verify if a paper is retracted or flagged on PubPeer, the screening provides verification routing to check status and prior concerns across databases like Retraction Watch, ORI, PubMed/Crossmark, and scite.

What types of files or inputs do I need to provide for biomedical forensic screening?

For biomedical forensic screening, you can provide a target paper or DOI, a figure image such as a microscopy panel, numeric summaries, or an .xlsx supplement to initiate the evidence-bound analysis.