Systemic institutional pattern analysis tool. Determines whether documented events represent isolated incidents or components of a systemic pattern of democratic or institutional erosion.
Built on the methodology developed for the Sweden OSINT Report 2026. Reusable on any country or institution.
Users paste verified source material — official statistics, news reporting, government documents — and the tool applies a structured analytical framework to identify cross-sectoral patterns, gaps between official narrative and primary data, and institutional response patterns.
Output is always structured in three explicit layers:
- Layer 1 — Documented facts (sourced only)
- Layer 2 — Observed patterns across those facts
- Layer 3 — Probability assessment
- Methodology & System Prompt full analytical framework
- Reference Implementation — Sweden 2019–2026 — first application of this framework to a real dataset, covering Swedish institutional patterns across a full mandate period
- Get a free Anthropic API key at console.anthropic.com
- Open the tool
- Paste your API key (stored in memory only, never logged)
- Select country/institution and time period
- Choose analytical dimensions
- Paste labeled source material
- Run analysis
Each analysis costs approximately $0.01–0.05 depending on source volume.
- Oversight bypass
- Institutional penetration
- Statistical misdirection
- Selective rule of law
- Parliamentary mechanics
- Surveillance & control
- Narrative vs. reality
- Elite impunity
This tool has no backend. It is a static React application that calls the Anthropic API directly from your browser. No data is collected, stored, or transmitted anywhere except from your browser to Anthropic's API using your own key.
- Go to: Pattern-Analyzer App
- Input Antrophic API-KEY.
- Choose area of interest.
- Input data to analyze.
- Run Pattern Analysis
- Read results
- Analysis is generated from the source material provided.
- All factual claims derive from that material.
- Pattern assessments and probability estimates are analytical outputs, not documented facts.
- Independent verification of all sources is recommended before any use of this analysis.
npm install
npm run dev- Fork this repository
- Go to vercel.com and sign in with GitHub
- Click "New Project" -> select this repository
- Click Deploy — no configuration needed
- Done. Vercel gives you a public URL automatically.
Every push to main deploys automatically.
- Fork this repository
- Go to netlify.com and sign in with GitHub
- Click "Add new site" -> "Import an existing project"
- Select this repository
- Build command:
npm run build - Publish directory:
dist - Click Deploy
The analytical framework encodes the following principles:
Refuse isolated incident framing — Events are tested for systemic significance regardless of sector or news cycle.
Formal vs. functional distinction — Institutional existence is separated from institutional performance.
Narrative vs. primary data — Government claims are cross-referenced against data from those same institutions' own agencies.
Response as data — How an institution responds after exposure is weighted equally with the incident itself.
Statistical decomposition — Aggregate statistics are always broken into subcategories before conclusions. Aggregate trends frequently conceal contradictory subcategory movements.
Three-layer output — Facts, patterns, and probability assessments are always distinct and explicitly labeled.
This tool was developed from the analytical methodology used in the Sweden: A Democracy Under Systemic Stress report (2025–2026), which applied this framework to Swedish institutional data between 2019 and 2026.
The tool is designed to be country-agnostic and reusable by researchers, journalists, and analysts working on institutional integrity in any democratic context.
If you apply this framework to another country or institution and want to add it as a reference implementation, open a pull request with your dataset in the docs/ directory.
MIT — use freely, modify freely, redistribute freely.