Swedish Political Intelligence Platform — democratic transparency, evidence-based analysis and AI-generated political news, powered by official open data and a fully autonomous agentic newsroom.
|
🕵️ Political intelligence · 🔍 Democratic transparency · 🤖 AI-generated news · 📊 50+ years of evidence Riksdagsmonitor monitors Sweden's Riksdag (Parliament), the Government (Regeringskansliet) and public agencies (Myndigheter) with structured intelligence techniques — ACH, SWOT, PESTLE, STRIDE, political-risk scoring and OSINT/INTOP tradecraft — applied to 349 current MPs, 2,494 historical politicians (1971–2024), 3.5M+ votes and 109,000+ parliamentary documents. An autonomous AI newsroom — 11 agentic workflows, Claude Opus, zero human editors — turns this evidence into publication-ready intelligence articles in 14 languages, every day. 
|
Strengthen Swedish democracy through systematic transparency.
Riksdagsmonitor exists to put rigorous, evidence-based political intelligence in the hands of every citizen, journalist, researcher and policymaker. We combine:
- 🕵️ Structured intelligence tradecraft — ACH, SWOT, PESTLE, STRIDE, political-risk scoring, OSINT/INTOP methodology, ICD-203 Key Judgments
- 🔍 Democratic transparency — every claim traceable to a primary source (
dok_id, vote count, named MP, government document) or it does not get published - 🤖 AI-generated political news — a fully autonomous newsroom turning open data into 14-language analysis articles, daily
- 📊 50+ years of historical evidence — 1971–2024 voting records, document corpus, and party-evolution analytics from the Citizen Intelligence Agency (CIA) platform
- ⚖️ Neutrality & GDPR by design — equal treatment of all 8 parliamentary parties, public-data only, privacy-by-design architecture, explicit democratic safeguards
The platform is non-partisan, open-source (Apache-2.0), and operated under the Hack23 ISMS with full ISO 27001:2022 / NIST CSF 2.0 / CIS Controls v8.1 alignment. It does not collect user data, does not run ads, does not push narratives, and is architecturally designed so it cannot be weaponised for partisan influence.
Five flagship pages anchor the public site. Each is multilingual (14 languages), WCAG 2.1 AA accessible, SEO-optimised with JSON-LD structured data and hreflang alternates, and CSP-hardened with Subresource Integrity (SRI) on every CDN asset.
| Icon | Page | What it does |
|---|---|---|
| 🌐 | Live Platform — riksdagsmonitor.com |
The public homepage and primary entry point. Front-loads the current coalition status (e.g. Tidö Agreement — 176/349 seats, fragility indicators, CIA risk alerts), then deep-dives into the full intelligence stack:
|
| 🕵️ | Political Intelligence Index |
The single canonical entry point for OSINT/INTOP researchers. Catalogues every methodology that governs analysis on the platform:
_sv, _de, _ar, …).
|
| 📰 | AI-Generated News & Analysis |
The world's first fully AI-driven political-intelligence newsroom for parliamentary monitoring. 11 specialised agentic workflows (10 scheduled + 1 on-demand, powered by Claude Opus via GitHub Copilot Coding Agent) autonomously produce daily intelligence articles:
NewsArticle.isBasedOn, links back to the analysis artifacts that produced it, and is published in 14 languages.
|
| 📊 | CIA Intelligence Dashboard |
Interactive Chart.js / D3.js intelligence dashboard powered by the Citizen Intelligence Agency (CIA) data products:
|
| 🗺️ | Sitemap (Human + Machine) | Human-readable navigation of every page on the platform plus the machine-readable sitemap.xml and rss.xml feeds. Daily refreshed; honours robots.txt and search-engine indexing best practices. |
🌍 14 supported languages: English · Swedish · Danish · Norwegian · Finnish · German · French · Spanish · Dutch · Arabic (RTL) · Hebrew (RTL) · Japanese · Korean · Chinese.
"While traditional newsrooms debate whether AI will replace journalists, Riksdagsmonitor already runs a fully autonomous political-intelligence newsroom — 11 agentic workflows, 14 languages, zero human editors, and a publication schedule that would bankrupt any legacy outlet trying to keep up."
Traditional AI-generated news is shallow — it rewrites press releases. Riksdagsmonitor's pipeline performs deep political-intelligence analysis before a single sentence of an article is written:
- 🔎 Source verification — every claim traced to official Riksdag/Regering data via the
riksdag-regeringMCP server (32+ tools) - ⚖️ Multi-party balance — algorithmic fairness across all 8 parliamentary parties, no editorial bias possible
- 📊 Quantitative rigor — voting cohesion metrics, attendance scoring, legislative productivity indices, ICD-203 Key Judgments, 45-rule transparency scoring
- 🌐 14-language reach — culturally adapted political analysis with RTL support (Arabic / Hebrew), not machine translation
- 🛡️ GDPR-compliant OSINT — only public parliamentary data; political opinions are processed under GDPR Art. 9(2)(e) (manifestly made public) / 9(2)(g) (substantial public interest)
- 🚦 Hard analysis gate — every news workflow MUST produce 9 (single-type) or 14 (Tier-C) analysis artifacts on disk before article generation; the gate at
.github/prompts/05-analysis-gate.mdis non-negotiable - 🧪 Quality gates — HTMLHint validation, link checking, accessibility (WCAG 2.1 AA) and translation parity in CI before publication
flowchart LR
A[📥 MCP + CIA + SCB + IMF<br/>data download] --> B[📐 Apply methodology<br/>analysis/methodologies]
B --> C[📋 Populate templates<br/>analysis/templates]
C --> D[📂 Write 9 or 14 artifacts<br/>analysis/daily/$DATE/$SUBFOLDER]
D --> E{🚦 Analysis Gate<br/>prompts/05}
E -- pass --> F[📰 Aggregate → article.md<br/>scripts/aggregate-analysis.ts]
E -- fail --> C
F --> G[🌐 Sanitised HTML<br/>scripts/render-articles.ts]
G --> H[🌍 Translate into<br/>12 other languages]
H --> I[🔀 One PR per article type<br/>S3 + CloudFront deploy]
style A fill:#0a0e27,stroke:#00d9ff,color:#e0e0e0
style E fill:#dc3545,stroke:#b02a37,color:#fff
style I fill:#1a1e3d,stroke:#ffbe0b,color:#e0e0e0
A new .md artifact written anywhere under analysis/daily/$DATE/$SUB/ is enough to publish an English + Swedish HTML article on the next CI build — there is no manual scaffolding, no template fill-in, and no per-type generator class. The dedicated news-translate workflow then propagates to the remaining 12 languages out-of-band. See Article-Generation.md for the full agentic-workflow contract.
| Time (UTC) | Workflow | Coverage | Frequency |
|---|---|---|---|
| 🌅 04:00 | Committee Reports | Utskottsbetänkanden analysis, voting breakdowns | Mon–Fri |
| 🌅 05:00 | Propositions | Government bills, legislative impact assessment | Mon–Fri |
| ☀️ 06:00 | Motions | Opposition proposals, party-strategy decoding | Mon–Fri |
| ❓ 07:00 | Interpellations | Ministerial accountability, evasion detection | Mon–Fri |
| 🔮 07:00 | Week Ahead | Parliamentary calendar preview, agenda intelligence | Friday |
| 📅 08:00 | Month Ahead | Strategic outlook, coalition forecasting | 1st of month |
| 🔍 10:00 & 14:00 (Mon–Fri); 12:00 (Sat/Sun) | Real-Time Monitor | Breaking political developments, flash analysis | Mon–Fri (×2) + weekends |
| 🌍 11:00 & 17:00 (Mon–Fri); 14:00 (Sat/Sun) | Translate | 12 additional languages from EN/SV cores | Daily |
| 🌆 18:00 (16:00 Sat) | Evening Analysis | Deep-dive intelligence synthesis | Mon–Sat |
| 📊 09:00 | Weekly Review | Week-in-review scorecard, party performance | Saturday |
| 📈 10:00 | Monthly Review | Comprehensive monthly intelligence assessment | 28th of month |
Authoritative schedules defined in
.github/workflows/news-*.lock.yml— see.github/workflows/README.mdfor the complete inventory.
Result: dozens of articles per week across 14 languages — hundreds of localised intelligence products each month, generated autonomously with zero editorial intervention, every one of them auditable down to its source dok_id.
Every analysis on the platform is governed by an explicit, version-controlled methodology. The full library is browseable at https://riksdagsmonitor.com/political-intelligence.html; the canonical sources live in analysis/.
| Layer | Document | What you'll find |
|---|---|---|
| 🧭 Framework | analysis/README.md |
Artifact taxonomy, 9-artifact / 14-artifact contract, on-disk daily layout |
| 📐 Methodology library | analysis/methodologies/README.md |
11 methodology documents — AI-Driven Analysis Guide · OSINT Tradecraft Standards · Political Risk Methodology · Political SWOT Framework · Political Threat Framework · Political Classification Guide · Electoral Domain Methodology · Synthesis Methodology · Per-Document & Per-Artifact Methodologies · Strategic Extensions (scenario / wildcard / long-horizon) · Structural Metadata · Political Style Guide |
| 📋 Template library | analysis/templates/README.md |
23 templates — 8 core single-type (T1–T8) plus 15 extended / Tier-C (executive-brief, scenario-analysis, coalition-mathematics, election-2026, historical-parallels, comparative-international, devil's-advocate, intel-assessment / ICD-203 Key Judgments, …) |
| 🚦 News-generation contract | .github/prompts/README.md |
8 bounded-context prompt modules + Tier-C extension; the single blocking analysis gate |
| ⚙️ Workflow orchestration | .github/workflows/README.md + WORKFLOWS.md |
How each news-*.md source compiles to a hardened .lock.yml with SHA-pinned actions, egress firewall (Squid + iptables) and five-layer safe-outputs |
| 🔍 IMF integration | analysis/imf/README.md + .github/aw/ECONOMIC_DATA_CONTRACT.md |
Macro/fiscal/monetary/external context with T+5 projections; canonical pattern for every economic claim |
| 🏛️ Statskontoret integration | analysis/statskontoret/README.md |
Swedish agency structure and central-government budget execution (årsutfall / månadsutfall / tidsserier) |
| 🤖 Specialist personas | .github/agents/README.md |
14 persona agents incl. intelligence-operative, news-journalist, content-generator + 9 workflow-specialist agents + shared developer.instructions.md |
| 🧠 Skills library | .github/skills/README.md |
91 skills across 12 functional categories — 11 political-intelligence, 5 journalism, 14 ISMS/security, 13 GitHub Agentic Workflows, … |
Every claim must tie to: a dok_id citation, a named actor, a vote count, or a primary-source URL. Generic statements without evidence are rejected by the analysis gate. AI is an accelerator — never an excuse for shallow output.
Minimum 2 complete iterations for every analysis. Pass 1 produces the artifact; Pass 2 reads everything back and improves every section (stronger evidence, deeper analytic rigor, broader stakeholder coverage, quantified risk). Workflows that complete in under 75 % of their allocated time are treated as failed. See AGENTS.md §AI-FIRST.
Five flagship Chart.js / D3.js dashboards on https://riksdagsmonitor.com/ (and the consolidated CIA Intelligence Dashboard):
| # | Dashboard | Coverage | Visualisations | Data |
|---|---|---|---|---|
| 1 | 🌡️ Seasonal Activity Patterns | 2002–2025 quarterly (23+ years) | Heat maps, time series, Z-score anomaly detection (|Z| ≥ 2.0) | cia-data/seasonal/ |
| 2 | 👤 Politician Dashboard | 349 current MPs | Top-10 rankings, 45-rule risk profiles, influence metrics | cia-data/politician/ |
| 3 | 🗳️ Pre-Election Monitoring | Q4 2023 → 2025 | Historical comparisons, election-year vs non-election Q4 patterns, early-warning indicator matrix | cia-data/pre-election/ |
| 4 | 🏛️ Party Performance & Effectiveness | 1990–2026 (37 years, 8 parties) | Effectiveness trends, comparative bars, coalition-alignment, momentum with P50/P90 percentiles | cia-data/party/ |
| 5 | 🚨 Anomaly Detection & Early Warning | 2002–2026 (41 quarters) | Timeline · Z-score distribution · type breakdown · severity heat map · recent-anomaly feed | cia-data/seasonal/ |
Dashboard properties — local-first data loading with 1-hour cache; WCAG 2.1 AA accessible (keyboard, screen reader, 4.5:1 contrast); 14-language; responsive 320 px → 1440 px+; CSP-compliant with SRI hashes (sha384) on all CDN resources.
Live numbers (updated daily at 03:00 CET via update-cia-csv-data.yml):
| Metric | Value | Note |
|---|---|---|
| 👥 Current MPs | 349 | All active Members of Parliament |
| 📜 Historical politicians | 2,494 | 1971–2024 (50+ years) |
| 🗳️ Votes analysed | 3.5+ million | Comprehensive voting-record corpus |
| 📄 Documents processed | 109,000+ | Parliamentary documents (motions, propositions, interpellations, …) |
| 🏛️ Committee documents | 8,740 | Committee work tracked |
| 2,308 | Across 45 transparency rules | |
| 🇸🇪 Political parties | 8 | All Riksdag-represented parties |
| ⏱️ CIA subsystems | 15 | anomaly · coalition · committee · distribution · election · election-cycle · ministry · parties · party · percentile · politician · pre-election · risk · seasonal · voting |
| 📰 News articles published | 2,669+ files under news/ |
14 languages |
Data source: extraction_summary_report.csv · cached in cia-data/production-stats.json (24 h freshness).
Riksdagsmonitor uses a provider-tiered data architecture, with each provider chosen for its area of strength.
| Tier | Provider | Scope | Access |
|---|---|---|---|
| 🏛️ Parliamentary primary | Riksdagen Open Data | Documents, motions, votes, MPs, speeches, committees | riksdag-regering MCP server (32+ tools) |
| 🏢 Government primary | Regeringskansliet | Propositions, SOU, Ds, directives, press releases | riksdag-regering MCP server |
| 📈 Primary economic | IMF (Datamapper REST + SDMX 3.0) | GDP, growth, unemployment, inflation, fiscal balance, debt, current account, bilateral trade, commodity prices, exchange rates, government spending by COFOG function — with T+5 projections | Pure-TypeScript client scripts/imf-client.ts (intentionally non-MCP) |
| 🇸🇪 Swedish ground truth | SCB (PxWeb v2) | Swedish monthly labour (AKU), monthly inflation (KPI), regional/municipal, budget execution | scb MCP server (@jarib/pxweb-mcp@2.0.0, 1,200+ tables) |
| 🏛️ Statskontoret | Statskontoret | Authority count, dept grouping, leadership form, FTE / headcount, central-government budget outturns | scripts/statskontoret-client.ts |
| 🌍 Non-economic residue | World Bank | Governance (WGI, source=75), environment, social/education residue, defence historicals |
world-bank MCP server (worldbank-mcp@1.0.1) |
| 🗳️ Election authority | Valmyndigheten | Election results, voter turnout, electoral statistics | Public datasets |
| 💰 ESV | Ekonomistyrningsverket | Government budget and spending data | Public datasets |
| 🕵️ Citizen Intelligence Agency | Hack23/cia | 15 CIA subsystems consumed nightly via update-cia-csv-data.yml |
JSON / CSV exports |
Why this split — IMF uses uniform SNA 2008 / GFSM 2014 / BPM6 methodology across countries (essential for cross-country comparison), publishes T+5 projections (essential for look-ahead workflows), and has fresher data than World Bank's economic indicators. World Bank remains canonical for the classes IMF does not publish (WGI governance, environment). SCB is the Swedish-specific ground-truth layer. Banned phrases (e.g. "the World Bank reports Swedish GDP growth of …") and vintage discipline (data > 6 months old → annotation required) are enforced by CI per .github/aw/ECONOMIC_DATA_CONTRACT.md v2.1.
Full system architecture (C4 Context / Container / Component / Dynamic views) lives in
ARCHITECTURE.md(v2.2).
- Frontend — Static HTML5 / CSS3 with TypeScript-built Chart.js / D3.js dashboards (no SPA framework, mobile-first, cyberpunk theme)
- Build — Vite 8 (ES modules, code splitting, SRI via
vite-plugin-sri-gen) - Visualisation — Chart.js 4 + D3.js 7, hosted locally on CloudFront
- Testing — Vitest 4 (2,890 unit tests, 100 % pass rate, 70 % line coverage) + Cypress 15 (E2E)
- Hosting — AWS CloudFront + S3 dual-region (us-east-1 primary, eu-west-1 replica) via OIDC; GitHub Pages as DR fallback
- CI/CD — 43 GitHub Actions workflows (21 standard
.yml+ 11 agentic.mdsources + 11 compiled.lock.yml); SHA-pinned,step-security/harden-runnereverywhere - Data Platform — Citizen Intelligence Agency (CIA) Java/Spring Boot backend + 15 CIA subsystems
- Runtime — Node.js 25.x
| Current State | Future State |
|---|---|
| 🏗️ Architecture | 🚀 Future Architecture |
| 📊 Data Model | 📊 Future Data Model |
| 🔄 Flowcharts | 🔄 Future Flowcharts |
| 🔄 State Diagrams | 🔄 Future State Diagrams |
| 🗺️ Mindmap | 🗺️ Future Mindmap |
| 💼 SWOT | 💼 Future SWOT |
Full controls in
SECURITY_ARCHITECTURE.md· threat model inTHREAT_MODEL.md· CRA conformity inCRA-ASSESSMENT.md.
Classification (per Hack23 Classification Framework)
| Dimension | Level | Note |
|---|---|---|
| 🔒 Confidentiality | 🟢 Public | All data intentionally disclosed (Swedish open data + website content) |
| ✅ Integrity | 🟠 High | Automated validation, GPG-signed commits, SLSA build provenance |
| 🟢 Availability | 🟠 High | 99.998 % design target (CloudFront 99.9 % SLA + multi-region S3 + GitHub Pages DR) |
| 🏷️ Privacy | 🟠 Personal (public officials only) | GDPR Art. 6(1)(e/f); Art. 9(2)(e/g) for political opinions; no end-user PII, no accounts, no ads, no tracking |
| ⏱️ RTO / RPO | 1–4 h / 4–24 h | Automated multi-region failover, daily data refresh |
| 💰 Business impact | 🟢 Low (financial) · 🟡 Moderate (reputational) | Open-source project, no revenue dependency |
- ISO 27001:2022 — 7 Annex A controls implemented · NIST CSF 2.0 — 6 functions aligned · CIS Controls v8.1 — 6 controls implemented
- GDPR — public-interest / legitimate-interest grounds for public-official data; political opinions under Art. 9(2)(e)/(g)
- EU CRA — self-assessment in
CRA-ASSESSMENT.md - OpenSSF Best Practices — Project #12069
- 🔒 HTTPS-only — TLS 1.3, HSTS, CSP, X-Frame-Options, X-Content-Type-Options
- 🧱 Five-layer agentic-workflow security — read-only agent tokens, zero secrets in agent context, containerised + Squid/iptables egress firewall, safe-outputs validation, AI threat-detection scan
- 🔐 SHA-pinned actions +
step-security/harden-runner+ Dependabot + CodeQL + Secret Scanning + dependency-review + OIDC-only AWS deploy - 📜 Aligned ISMS policies — Information Security Policy · Secure Development Policy · Threat Modeling · AI Policy · Vulnerability Management · Change Management · Incident Response Plan · Open Source Policy
| Metric | Status |
|---|---|
| Risk level | 🟢 LOW (5.52 / 10.0 — 99.7 % risk reduction) |
| HTML validation | ✅ 0 errors (HTMLHint) |
| Dependencies | ✅ Dependabot clean |
| Secrets | ✅ Secret Scanning enabled |
| Code scanning | ✅ CodeQL active |
The platform's reusable shared utilities are published as riksdagsmonitor with SLSA build provenance.
npm install riksdagsmonitorIncludes — Theme System (cyberpunk, WCAG AA) · Chart Factory (Chart.js with responsive breakpoints + keyboard nav) · Resilient Data Loader (retry, cache, CSV/JSON) · DOM utilities · Full TypeScript types · 12 dashboard modules · CIA intelligence modules.
import { getActiveThemeColors, BREAKPOINTS, getPartyColor } from 'riksdagsmonitor';
import { loadJSON, loadCSV, createDataSource } from 'riksdagsmonitor';
import { showLoadingState, formatNumber, debounce } from 'riksdagsmonitor';
import { createChart, initDashboardSection } from 'riksdagsmonitor/shared/chart-factory';
import { CIADataLoader } from 'riksdagsmonitor/cia/data-loader';
import { CIADashboardRenderer } from 'riksdagsmonitor/cia/visualizations';Peer dependencies — chart.js d3 papaparse (required for dashboards); chartjs-plugin-annotation (optional, loaded conditionally).
Node.js ≥ 25 · npm ≥ 10 · Git with GPG signing · GitHub MFA + SSH keys.
git clone git@github.com:Hack23/riksdagsmonitor.git
cd riksdagsmonitor
npm install
npm run dev # Vite dev server with hot reload → http://localhost:8080
# Tests
npm test # Vitest unit (2,890 tests)
npm run test:coverage # with coverage
npm run cypress:open # E2E interactive
npm run e2e # full E2E (build + preview + Cypress)
# Quality
npm run htmlhint # HTML5 validation
npm run linkcheck # link integrity (linkinator)
# Production
npm run build # Vite production build → dist/
npm run preview # http://localhost:4173- HTMLHint validation · Linkinator · Vitest (2,890 tests) · Cypress · Vite build · Dependency review · CodeQL · Secret scanning · Translation validation
- Releases — workflow_dispatch or
v*.*.*tag → SBOM (SPDX) + SHA-256 + SLSA Build Provenance attestations + dual deployment (S3/CloudFront primary, GitHub Pages DR) gh attestation verify riksdagsmonitor-vX.Y.Z.zip -R Hack23/riksdagsmonitor
See WORKFLOWS.md and RELEASE_PROCESS.md for the canonical reference.
Riksdagsmonitor uses GitHub Copilot personas, skills and agentic workflows as first-class automation. Directory READMEs are the single source of truth; AGENTS.md and SKILLS.md are the long-form catalogs.
| Surface | Catalog | Count |
|---|---|---|
| 🤖 Custom agents | .github/agents/README.md |
24 files (14 personas + 9 workflow-specialists + 1 shared developer-instructions) |
| 🧠 Skills | .github/skills/README.md |
91 skills across 12 functional categories |
| 📜 Prompt modules | .github/prompts/README.md |
8 bounded-context modules + Tier-C extension |
| ⚙️ Workflows | .github/workflows/README.md |
43 files (21 standard + 11 agentic sources + 11 compiled) |
| 🔌 MCP servers | .github/copilot-mcp.json |
8 — riksdag-regering, scb, world-bank, github (insiders), filesystem, memory, sequential-thinking, playwright |
14 persona agents (assignable via assign_copilot_to_issue):
security-architect · documentation-architect · quality-engineer · frontend-specialist · isms-compliance-manager · deployment-specialist · devops-engineer · intelligence-operative · news-journalist · content-generator · data-pipeline-specialist · data-visualization-specialist · task-agent · ui-enhancement-specialist.
9 workflow-specialist agents (.agent.md, invoked from workflows):
agentic-workflows · ci-cleaner · contribution-checker · create-safe-output-type · custom-engine-implementation · grumpy-reviewer · interactive-agent-designer · technical-doc-writer · w3c-specification-writer.
Article-Generation.md— Canonical agentic-newsroom contract (data → analysis → article → HTML → S3)AGENTS.md·SKILLS.md— Copilot persona & skill catalogsWORKFLOWS.md— End-to-end CI/CD reference (v7.2)RELEASE_PROCESS.md— Release with attestations & docs-as-codeTRANSLATION_GUIDE.md— 14-language standards & glossarySECURITY.md·CONTRIBUTING.md·CODE_OF_CONDUCT.md·LICENSE
FUTURE_WORKFLOWS.md·FUTURE_MINDMAP.md·FUTURE_ARCHITECTURE.md·FUTURE_SECURITY_ARCHITECTURE.md·FUTURE_SWOT.md
- 📚 API Documentation (JSDoc) · Test Coverage · Cypress E2E Reports
- 🕵️ CIA Platform · CIA JSON Export Specs
- 🛡️ Hack23 ISMS-PUBLIC · 📰 Hack23 Blog
From agentic news generation to AGI-powered democratic intelligence — an 11-year evolution.
Detailed planning:
FUTURE_WORKFLOWS.md·FUTURE_MINDMAP.md·FUTURE_ARCHITECTURE.md
%%{init: {"theme": "dark", "themeVariables": {"primaryColor": "#00d9ff", "primaryTextColor": "#e0e0e0", "primaryBorderColor": "#ff006e", "lineColor": "#ffbe0b", "secondaryColor": "#1a1e3d", "tertiaryColor": "#0a0e27"}}}%%
timeline
title Riksdagsmonitor Evolution — 2026 → 2037
section Phase 3 — Foundation (2026)
Q1-Q2 : ✅ TypeScript migration · 11 agentic news workflows
: ✅ 14 languages · S3 + GitHub Pages dual deploy
Q3-Q4 : 🔄 CIA pipeline integration · predictive dashboards
section Phase 4 — Predictive Intelligence (2027–2029)
2027 : Real-time streams · ML election models · Nordic expansion (DK · NO · FI)
2028 : Multi-modal (audio/video) · AI editorial board · self-healing pipelines
2029 : Fully autonomous · expert-level analysis · API platform launch
section Phase 5 — Scale (2030–2033)
2030-2031 : EU-wide coverage (27 + EP) · sovereign-AI evaluation
2032-2033 : Global parliamentary watch (100+ democracies) · every UN language · pre-AGI
section Phase 6 — AGI Era (2034–2037)
2034-2035 : AGI-enhanced intelligence · 195 parliaments · predictive governance
2036-2037 : Transformative democracy platform · democratic safeguards architecture
| Year | Automation assets | AI model | Key capability |
|---|---|---|---|
| 2026 | 44 → 50 | Opus 4.7–4.9 | 🤖 Agentic news generation (current) |
| 2027 | 50–55 | Opus 5.x | 🔮 Predictive analytics & Nordic expansion |
| 2028 | 55–65 | Opus 6.x | 🎙️ Multi-modal content |
| 2029 | 65–75 | Opus 7.x | 🚀 Fully autonomous pipeline |
| 2030–2033 | 75–100 | Opus 8–10.x / pre-AGI | 🌍 EU → Global coverage |
| 2034–2037 | 100–120+ | AGI / post-AGI | ⚡ Transformative democracy platform |
Version numbers are illustrative — actual products and paradigm shifts (quantum AI, neuromorphic computing) will vary. Architecture is designed for graceful adaptation while preserving democratic safeguards: human oversight maintained regardless of AI capability, anti-weaponisation by design, constitutional alignment encoded in platform architecture.
Contributions welcome under Hack23's secure-development standards.
- Fork the repository and create a descriptive feature branch
- GPG-sign every commit · enable MFA on your GitHub account
- Run quality checks locally (
npm run htmlhint && npm test && npm run build) - Submit a pull request with comprehensive description; address review feedback
- Never introduce security vulnerabilities; follow
CONTRIBUTING.mdandCODE_OF_CONDUCT.md
Hack23 AB (Org.nr 559534-7807) — Swedish cybersecurity and open-source intelligence consultancy.
- 🌐 www.hack23.com · 📰 Blog
- 💼 LinkedIn — Hack23
- 👨💻 Founder: James Pether Sörling, CISSP, CISM
- 🛡️ Public ISMS · 🕵️ CIA platform
Copyright © 2008–2026 Hack23 AB (Org.nr 559534-7807). Licensed under the Apache License 2.0 — see LICENSE.
🗳️ Empower citizens · 🔍 Strengthen democratic accountability · 🕵️ Illuminate the political process