261
Investigation
pages
35
Canonical
figures
119+
Master index
links
6/6
GC datasets
synced
0
Anonymous
sources

Data Sources

Official Government Records Only

Every claim on TENET5 is sourced from official government records: Health Canada annual reports, Statistics Canada databases (IMDB, CCHS, CPI), Auditor General performance audits, Federal Court and Supreme Court decisions, parliamentary Hansard, Ethics Commissioner findings, IRCC immigration data, DND defence reports, CMHC housing data, CRA tax statistics, OECD comparative data, NATO defence expenditure reports, and corporate SEC filings. Zero anonymous sources. Zero unverified claims.

Peer-Reviewed Academic Research

The Pearce Study (Western University, OMEGA — Journal of Death and Dying, Feb 2025, DOI: 10.1177/00302228251323299) provides peer-reviewed academic validation for the MAID cost-avoidance model. This is published research in a Sage Journals publication, not opinion or conjecture.

Infrastructure

Canonical Figure System

35 key statistics are maintained in a single source of truth (data/canonical_figures.json). A CI validator (an automated figure validator) scans all pages for discrepancies. An auto-injector (js/figures.js) ensures pages display current values. This prevents the kind of figure drift that undermines credibility.

GC Open Data Sync

An automated sync job pulls fresh data from Government of Canada open data portals: 6 proactive disclosure datasets, fiscal data (Budget 2025), lobbying registry updates, and Elections Canada data. This ensures the investigation uses current official data, not stale snapshots.

404 Error Reporter

Every page loads js/error-reporter.js which HEAD-checks all internal links and logs any 404s to localStorage. Broken links are highlighted with red outlines. The CEO can check errors via browser console: TENET5_ERRORS.printLog().

LIRIL AI Voice Narration

Every page includes data-narrate attributes that LIRIL reads aloud using Microsoft Hazel (British female voice). The narration system (js/liril-walkthrough.js) sequences sections and provides subtitle display. Voice selection is handled by js/liril-voice.js with an explicit target voice list and expanded male blocklist.

Accountability Analysis Methodology

Step 1 — Build the dossier

For each accountability axis (e.g. opioid_crisis, winnipeg_lab_nml), a JSON dossier is built from primary sources only — Auditor General reports, parliamentary committee transcripts, federal court judgments, public-health quarterly reports, lobbying registry exports, Hansard. Each dossier lists the executive decision-makers whose authority touched the file, with date ranges and statutory references.

Step 2 — Define the marking criterion

The marking criterion is the investigative claim made precise. For winnipeg_lab_nml, the criterion marks any actor whose tenure overlapped both the July 2019 NML escort-out and the unredacted-document fight. The marking function is auditable, deterministic, and committed to source. There is no "model" doing the marking — the criterion is a boolean over public-record fields.

Step 3 — Rank the marked actors

Each axis produces a concentration score: how many times more often a marked actor surfaces relative to a uniform baseline over the same public-record set. The score is computed deterministically from the dossier fields — no model, no inference. Five of fourteen axes resolve to a single dominant actor (score 1.000): arms exports, foreign interference, veterans affairs, CSIS oversight, and Winnipeg laboratory NML. The opioid-crisis axis tops the ladder at 6.37×.

Step 4 — Cross-axis intersection

Once each axis is ranked independently, the cross-axis pass identifies actors marked in multiple dossiers. Justin Trudeau is currently octuple-marked (8 of 8 covered axes; cross-axis score 24). Marco Mendicino, Dominic LeBlanc are triple-marked. Bill Blair, Harjit Sajjan, David Lametti, Jody Wilson-Raybould, David Vigneault are dual-marked. The intersection finding is the structural claim — that capture is concentrated, not distributed.

Step 5 — publish a SHA-256 hash of every file

Every dossier file, every ranking result, every cross-axis output is committed to a a SHA-256 hash via the published verification pipeline. Anyone can re-run it from the public source and confirm the same root. The current site root is published in data/merkle_verification_receipt.json. Reproduce, don't trust.

Where to read it

Investigation: accountability axes index · 10 ready-to-send citizen campaigns: .

Principles

1. Official Sources Only

No anonymous sources. No leaked documents. No unverified claims. Every fact is traceable to a specific official document. If a statistic cannot be verified from the cited source, it should not be on this site.

2. Independent Verifiability

Every page includes a "Sources" section with specific document references. Readers can verify every claim without trusting TENET5. The investigation is designed so that a sceptical reader can fact-check everything independently.

3. Open Source

The entire codebase — every HTML page, every JavaScript file, every tool — is publicly available at github.com/TENET-5. Every commit is visible. Every change is tracked. The investigation process itself is transparent.

4. Non-Partisan

TENET5 does not endorse parties, candidates, or policies. The investigation documents institutional performance from official data. Institutional capture is structural, not partisan — documented failures span multiple governments and parties.

Full investigation: Master Index (119+ links) → | Reading Order → | AI Research Methodology → | Press Kit → | FAQ →