A system that compiles embodied dynamics into justified N'Ko statements with cryptographic provenance. Every inscription is traceable to its source evidence through a typed IR pipeline.
Dual-Equilibrium Latent Learning - neural architecture operating on two timescales
Deterministic context slicing engine for bounded, reproducible slices
10 claim detectors converting z-trajectory to typed IR to N'Ko surface
Strongly-typed intermediate representation for each claim type
Proto-basin → Basin → Split/Merge/Retire state machine
Full API documentation for integration and development
The system does not predict what will happen; it detects when dynamics become constrained enough that a claim is warranted.
Old inscriptions are never rewritten. When the lexicon evolves, original inscriptions remain untouched.
Every claim maintains a chain: z-trajectory → ClaimDetector → Typed IR → Lexicon → N'Ko Surface → Proof Scaffold.
Every operation must be deterministic. Same inputs must produce byte-identical outputs across runs, sessions, and machines.
Every inscription follows this structure:
⟨operator-sigil⟩ ⟨time-marker⟩ : ⟨claim-body⟩ ; ⟨slots⟩
Example:
ߛ ⟦100.0–200.0⟧ : z(σ) ↓ ; ⟦home⟧ ; c=0.85
Reads as: "Stabilization claim from t=100.0 to t=200.0,
dispersion decreased, at place 'home', confidence 0.85"