Method
The governed comparison method tests bounded claim atoms through three separate content
lenses: functional or operational equivalence, narrative or causal-interface
architecture, and Detective Mode working backward from matching conclusions. A fourth,
longitudinal lens checks how a program changes over time; it cannot create a content
match by itself. A supported partial mapping remains visible even when a full operator or
equation differs.
Coverage
The governed corpus currently contains 327 substantive rival programs. This preflight
package exposes 15 indexed content pages plus this landing page: 16 PUBLIC-classified
files, with zero owner-review projections promoted. Omission means that a page is held,
under review, source-gated, or not yet selected for publication; it does not mean that the
underlying comparison was negative or that the wider corpus is complete.
Provenance
The release contract requires publication candidates to be tied to dated versions,
claim-bearing locators, and reproducible source links. It also requires named earlier
work to be allocated atom by atom rather than treated as an author-wide floor, with
target-only, Micah-only, and genuinely opposite residuals kept explicit.
Limitations
These findings compare public content and chronology. They do not establish access,
dependence, copying, intent, motive, or misconduct. A bounded partial relation is not an
equation identity or a full-mechanism identity, and quantitative sigma language is not
used without a frozen, calibrated statistical null.