1. Scope and system definition
Each study begins by defining the organization, product, service, technical boundary, research question, and time period. Ambiguous scope is recorded as a limitation rather than silently expanded.
2. Evidence hierarchy
Official technical documentation, standards, audited disclosures, public data, and direct system behavior receive priority. Independent reporting and expert analysis can add context. Anonymous claims and personal observations may guide inquiry but are not treated as confirmed facts.
- Claim and exact supporting evidence
- Source URL, publisher, and date
- Access and verification date
- Confidence level and verification status
- Recommended action linked to the finding
3. Efficiency assessment
The analysis considers architecture, operations, resource use, reliability, user impact, societal contribution, and improvement capacity. Numerical scores require defined inputs, units, weighting, and a reproducible calculation. Otherwise the report displays ‘Insufficient Data for Quantitative Assessment.’
4. Paradigm-neutral scrutiny
Institutional status, market power, tradition, and ideological familiarity do not exempt a subject from inquiry. Companies, brands, learning systems, philosophical positions, scientific tests, and doctrines are evaluated by disclosed evidence and consequences.
Criticism identifies assumptions, failure modes, and opportunities for redesign. It is not treated as hostility, and criticism alone is never evidence of wrongdoing.
5. Editorial approval
AI may retrieve sources, organize evidence, and draft analysis, but it cannot publish. An administrator reviews every source and claim before publication. Significant later changes create a new revision and require approval.
6. Corrections and responses
Researched organizations and affected people may submit evidence, corrections, or a formal response. A claim is marked resolved only after verifiable evidence or administrator confirmation—not solely because an organization states that a fix exists.