---
type: Concept
title: "Verify, don't trust (Maximum · General · Verifiable)"
description: "Replace trust with verification: expose what information was used, which permissions existed, which objective was optimized, which agent acted, what happened, who was contacted, why, under which rule, and the result."
resource: https://trooids.com/#verify
tags: [verifiability, transparency, mgv]
generated: { by: claude-fable-5-1/2026-09-22, at: 2026-09-22T09:45:11Z }
status: stable
license: CC-BY-4.0
sources:
  - id: word-brief-20260922
    resource: /knowledge/sources.md#word-brief
    title: "TrooIDs and Landing page generator — owner's ChatGPT conversation export, 2026-09-22 (20 missing propositions and the homepage hierarchy)"
    author: human:hbkerkhof
    last_modified: 2026-09-22T00:00:00Z
  - id: site
    resource: https://trooids.com/
    title: trooids.com (the page these concepts summarise)
    author: process:trooids-site-build-20260922
    last_modified: 2026-09-22T00:00:00Z
---
# Verify, don't trust
**Don't trust TrooIDs. Don't trust the AI. Don't trust the algorithm. Verify them.**[^word-brief-20260922]

The system should expose, where technically feasible and appropriate: what information was used · what permissions existed · what objective was being optimized · what application/agent acted · what action occurred · what external party was contacted · why · under which rule · what the result was. Principle: **Maximum · General · Verifiable (MGV)**.

Applied to this website: the page [How this site was built](https://trooids.com/built) lists every technical claim with its evidence class.

[^word-brief-20260922]: Owner's brief, 2026-09-22, proposition 4.