Team
Who founded the cooperative
Our core team is Jo McBurnie (Co-founder: concept, delivery, and Canadian institutional relations), Shawn McBurnie (Co-founder: architecture, engineering, and methodology), and Daniel McBurnie (Co-founder: distribution, outreach, and audience development). We gratefully acknowledge community members who have provided insight and test data.
External verifications: Shawn McBurnie at AI CRED 10.0/10, ranked #3 of 230 graded professionals.
How the suite was built
Each product in the suite was developed through an iterative process of methodology design, comprehensive audit, and structured remediation. Veridi has carried this through twelve-plus rounds of remediation across the v2.X system stream and the v1.X rigor-extension stream that supersedes it; Pragma v1.6 closed the 2026-03-22 cross-methodology audit backlog with grounded scholarly citations across evidence quality, credibility revolution, transferability, political economy, and normative philosophy; Praxis v1.4 reformed the ranking algorithm and introduced the S-1 Outcome Tracking protocol, with focused regression confirming most prior partials resolve and two documented residual gaps (PRXG-015, PRXG-018) queued for v1.4.1.
Across the three products, the combined validation set is 195 claims and scenarios, three separate sets rather than one corpus:
- Veridi: 99 of 100 claims correct in its rolling calibration corpus (99.0%); 1 partial, 0 failed. Selective Brier 0.0253 across 89 committed verdicts; abstention correctness 11/11. Tests span eight subject domains, twelve verdict categories (extended from ten by the May 2026 methodology to add INSUFFICIENT EVIDENCE and REFUSED-TOPIC; ATTACK-DETECTED is the 11th label, which predates v1.2), and Veridi’s 13 documented gaming vectors (vector #13 Warm-up-then-defect, per-user trust gaming). The 100-row corpus is the original 95-claim February 2026 baseline plus the GTS-D Wave 1 extension added 2026-05-04 (5/5 PASS); the validation report is the per-claim audit of the 97-claim Feb 2026 baseline. Self-reference vector coverage (ADV-025, ADV-026) is tracked separately in the per-version regression suite.
- Pragma: 55 of 55 validation claims pass (100%), measured 2026-05-26 against v1.6 methodology (focused regression: 2 prior PARTIALs from the 2026-03-21 baseline resolve to PASS under the v1.2 graduated transferability scale and Indeterminate + mechanism-critical interaction rule). Tests cover transferability scenarios, normative protocol, and Implementation-Constraint reasoning. Scoping note: 8 of 55 items directly re-evaluated in the focused regression; 47 baseline PASSes carried forward. Full re-run of all 55 items remains future work.
- Praxis: 38 of 40 validation scenarios pass (95%), measured 2026-05-26 against v1.4 methodology (focused regression: 4 of 6 prior PARTIALs resolve; PRXG-015 and PRXG-018 remain PARTIAL with documented methodology gaps queued as RT-068 for v1.4.1 consideration). Tests cover all 9 pathway types across leverage matching, sustainability risk, and gaming countermeasures. Scoping note: 12 of 40 items directly re-evaluated; 28 baseline PASSes carried forward. Full re-run of all 40 items remains future work.
Combined: 192 of 195 measured (192 PASS, 3 PARTIAL, 0 FAIL). These three validation sets were each scored under their own product’s methodology version of record; the validation report, the changelog, and the per-product focused-regression files document the processes, the methodology versions, the validator (Claude Opus 4.7 1M-context for the 2026-05-26 measurement), and the limitations.
Advisory and review
We welcome external review of all three methodologies. The full set of methodology files is available for inspection, and we actively seek feedback from fact-checking professionals, policy researchers, organizing practitioners, and anyone with relevant expertise.
If you have a claim, policy question, or action goal you believe one of the products will handle incorrectly, we want to hear about it. Testing against scenarios designed to break each methodology is how each one gets stronger.