Demo Window --d --h --m --s
Kenneth Vanini  ·  Sanctuary of Ma'at Research Institute

GUARDIAN

Three agents. One verdict.
The paper is weighed. No exception.

𓂀 𓋴 𓂀
Examine
$28B

Estimated annual cost of irreproducible research to the global economy. Behind that number: fabricated datasets, p-hacked statistics, forged ethics approvals, and citation manipulation — errors that propagate through literature for decades before anyone looks closely enough to weigh them.

Diederik Stapel
Social psychology. Fabricated data across 15 years of research.
58 retractions
Joachim Boldt
Anesthesiology. Forged ethics committee approvals.
90+ retractions
Surgisphere
Unverifiable proprietary database used in Lancet & NEJM COVID studies.
2 landmark retractions
Brian Wansink
Food psychology. Admitted to systematic p-hacking across research group.
15 retractions
𓁹
Statistical
Identical SDs, p-hacking, impossible effect sizes, fabricated variance
𓏛
Citation
Unsupported claims, self-citation loops, cherry-picked literature
𓆣
Methodology
Ethics approval, data availability, preregistration, COI
𓂀
Orchestrator
Synthesizes all findings. Renders the final verdict.

Guardian runs three specialist agents simultaneously against every manuscript, each calibrated to the fraud signatures that defined history's most damaging retractions. The Orchestrator weighs all findings and issues a single verdict: Critical, Flagged, or Compliant — with full evidence. Built on Complete.dev's collaborative multi-agent platform.

CRITICAL

Do not publish. The manuscript exhibits patterns directly matching known retraction cases. One or more findings represent a fundamental integrity failure — fabricated data, forged approvals, or an unverifiable database.

FLAGGED

Major revision required. Significant concerns detected — p-value clustering, unsupported absolute claims, missing ethics documentation — that must be addressed and verified before peer review.

COMPLIANT

No major fraud indicators detected. The manuscript passes all three agent checks. Standard peer review may proceed. Guardian is a tool, not a replacement for human expert judgment.

Kenneth Vanini is a specialist in ancient chronology and Ma'atic theology, and the founder of the Sanctuary of Ma'at Research Institute.

His research methodology — Systemic Inversion Methodology (SIM) — treats persistent, unchallenged narratives as variables to be tested. Guardian applies that same principle to academic science: when a paper's statistics never fail, the statistics themselves become the variable to examine.

Guardian was built for the Complete AI Hackathon 2026 on the Complete.dev collaborative multi-agent platform.

Methodology
Systemic Inversion Methodology
Platform
Complete.dev Multi-Agent Workspace
Domain
Research Integrity & Fraud Detection
Institute

"In the Hall of Two Truths, the heart is weighed against the feather of Ma'at. No exceptions for wealth, prestige, or the number of citations."

The principle applied to science