All of these are tools for structural analysis and do not replace the methods of each field; engineering-level systems still await empirical validation. Each tool belongs to one of the first problem classes: the science of generation delivers results as “problem–solution packages” — starting from a problem, registering predictions and thresholds in advance, publishing tools, code and measured results (failures included), with a certificate of verification.
Tools and platforms
Software
Rotor Whole-Genome Analyzer
Rotor analysis of gene regulatory networks: find cycles, assign phases, classify, enumerate attractors and basins. It has reproduced the Li 2004 yeast cell-cycle network with four independent checks, and is the basis of the first solution package (cell fate). Engineering
Platform
Xuanji Chengshi v0.4.1
Identifies the rotor that realises a wish, gives a three-valued existence verdict, designs ignition and restraint, and confirms closure by a withdrawal test. Prospective data are not analysed until 100 records have accumulated. No promise of certain success. Engineering
Benchmark
Dual-Track Evaluation Benchmark (DTEB)
Evaluates the conceptual structure of language models with dual-track concept generation theory. The seven DualTrack volumes of measurements come first (including predictions refuted by real data); the construct validity study follows a pre-registration. Programmatic
Database
Concept Generation Map (CGM)
A database of concept-generation relations and of circular structures; with the bisimulation identity criterion (Theorem) it serves the software-cycles and identity problem.
Model
Five-Phase monetary dynamics model v3.0
A Five-Phase model of money and socio-economic dynamics, run only as pre-registered predictions with a public ledger. Engineering Not investment advice.
Verification infrastructure
Data
Public claim base
Every public claim: permanent numbers, the nine-tier epistemic scale, dependencies, corrections and the tier change log; queryable by AI agents through REST and MCP.
Formalisation
Lean formalisation library (in progress)
Theorem-level results written one by one into the Lean 4 proof assistant: the pilot is overcoming = generation², followed by 5 as the smallest double-Hamiltonian closure, 5 ∤ 2ᵏ, solvability of the rotor-cell semigroup and soundness and completeness of RSC₀.
Ledger
Pre-registration and results ledger (in progress)
Predictions and thresholds of each solution package are registered before the data is seen; every result, PASS, PARTIAL or FAIL, goes into the public ledger.