High-dimensional mathematical analysis for AI agents. Built on 16-dimensional sedenion algebra — the structures dismissed as pathological — now formally verified in Lean 4 and deployed as a Model Context Protocol server.
CAILculator treats structures that conventional mathematics discards as computational resources. Every claim is machine-verified.
Non-zero elements P, Q where P × Q = 0. In 16D sedenion space there are 84 base patterns — formally enumerated and verified. CAILculator's oracle layer dispatches on (P,Q) object pairs, not labels.
✓ Lean Verified38 orders of magnitude dynamic range. Stability constant M = 2·(‖P‖²+‖Q‖²)/(α·e) is Lean-proved — not empirically estimated. Transform outputs are checked against the verified bound at runtime.
✓ Stability ProvedSix framework-independent bilateral zero-divisor patterns persisting from 16D through 256D in both Cayley-Dickson and Clifford algebras. Identified as lying on the E8 first shell, forming a single Weyl orbit.
✓ E8 Orbit VerifiedNative support for both non-associative Cayley-Dickson algebras and associative Clifford (geometric) algebras. The bilateral oracle runs parity-verified across both frameworks simultaneously.
✓ Clifford Bridge VerifiedZero Divisor Transmission Protocol — transmit data through verified mathematical gateways (S1–S6) to 32D and 64D spaces. Bilateral annihilation confirmed universal across all 50 Riemann zeros tested.
✓ Bilateral Collapse ProvedUniversal algebraic layer stays Lean-anchored. Domain projections are pluggable profiles: Journalism, Quant Equity, and RHI ship as reference implementations. Custom profiles available on request.
✓ v2.0 ArchitectureEvery tool dispatches on Lean-verified (P,Q) object pairs. No label-based heuristics. Stability bounds checked at output time.
The verified core stays universal. Domain projections are declared profile objects with versioned manifests, explicit evidence references, and auditable empirical dependencies.
Algebraic pattern detection optimized for data reporting. Treat structural collapses in budgets, consensus, and policy the way zero divisors treat dimensionality.
Dual-method analysis bridging HMM statistical baselines with algebraic structural analysis. OHLCV coefficient mapping to sedenion gateways, RSI/MACD/Bollinger interpretation.
Spectral research mapping and prime embedding analysis. The active profile powering an open-science Riemann Hypothesis investigation, with Phase 72 introducing the Sedenionic Hamiltonian H(s).
Core algebraic properties are machine-verified in Lean 4 with Aristotle co-authorship. Every oracle tool has non-triviality tests confirming it produces distinct outputs on distinct inputs.
Proves P·Q = 0 ∧ Q·P = 0 for all Canonical Six pairs. The oracle used to gate every ZDTP transmission.
Stability constant M, unconditional absolute convergence, pattern invariance across dimensional scaling.
Canonical Six confirmed on E8 first shell, forming a single Weyl orbit generating a 24-element zero divisor family.
API key delivered within 24 hours of subscription. Students save with code STUDENT2025 at checkout.
STUDENT2025 at checkout for savings off Individual pricing.
Python 3.10–3.13, 64-bit. Windows 10/11, macOS 10.15+, Ubuntu 20.04+.
Open %APPDATA%\Claude\claude_desktop_config.json and add:
Open ~/Library/Application Support/Claude/claude_desktop_config.json and add:
CLAUDE CODE
GEMINI CLI