Methodology
What each check does
For each of the three theory groups, the suite:
- Runs the physics — either the already-existing, previously-validated Python implementation (Group A’s package; Group B’s real 45-point orbit gamma-scan), or a new direct implementation of the equations as written in the book where no code existed (all of Group C).
- Extracts a measurable quantity from the run — e.g. whether the orbit stays bounded, whether the Kretschmann scalar is finite at r=0, whether ISCO shift scales as g³.
- Compares it against the book’s stated claim, with an explicit tolerance, and reports PASS / FAIL / NOT_TESTABLE / SKIPPED / ERROR.
Scope
This validates the load-bearing quantitative claims of each theory group — the numbers a physicist would actually check — not a mechanical simulation of all ~250 chapters across the books, most of which are conceptual or philosophical with no associated number.
Toy systems, named as such
The Ξ-selection and structural-stability demonstrations use an illustrative toy dynamical system. They show the claimed qualitative behavior — structured, stable, generative worlds beat trivial/frozen or divergent/chaotic ones under Ξ-maximization — they are not a claim that this toy system literally is the pre-physical selection process.
What’s deliberately absent
SPARC galaxy rotation-curve validation is intentionally absent: the real survey data isn’t present on the machine that produced this corpus — a known, named gap, not something worked around with a substitute. See Validation → SPARC.
Tolerance
Where a claim’s expected value is an order-of-magnitude estimate in the book (e.g. “γc ~ 0.03”), the check uses a correspondingly loose tolerance — it verifies the scaling/order of magnitude holds, not that the toy simulation reproduces a full high-resolution validated run to high precision.