Methodology
Complete transparency about how this site collects, verifies, scores, and presents information. Every methodology described here is implemented in the codebase and can be independently verified.
Our Core Standard
This site is built on verifiable facts — sourced from wire services (AP, Reuters), fact-checkers (PolitiFact, Snopes, FactCheck.org), official records, court documents, and video evidence (C-SPAN, PBS). Opinion content is the exception, not the rule, and is always clearly labeled.
What Is an Incident?
An incident is any publicly documented action, statement, lie, policy decision, scandal, or event involving the Trump presidency that is sourced from major news outlets, fact-checkers, or official records. Each incident is categorized by type, tagged by topic, rated for severity on a 1–10 scale across multiple dimensions (presidential norms, adult behavior, global reputation impact), and linked to primary sources.
Status Workflow
Every incident passes through a documented workflow: Draft (scanner-discovered or manually created) → Pending Review (flagged for additional review) → Published (approved and visible on the site). Incidents can also be Rejected (not worth publishing), Archived (previously published, now hidden), or Not Relevant (AI-assessed relevance score below threshold). Only published incidents appear on public pages.
Data Storage
All data is stored in Azure Cosmos DB (serverless, NoSQL) across 10 dedicated containers — incidents, lies, patterns, scanner logs, page views, messages, social media posts, lie comparisons, site settings, and diagrams. There are no runtime seed data files. The database is the single source of truth.