🪨 guideStone

The Verification Class: How Computation Becomes Proof

🪨✅ guideStone is a verification class — a quality grade for any ♻️📦 ecoBin in the ecosystem that certifies the artifact produces reproducible, self-proving results.

It is not a binary type. The binary ladder ( 1️⃣📦 UniBin → ♻️📦 ecoBin → 🧬📦 genomeBin) describes structure. It is not a deployment class. 🐭📡 fieldMouse describes where things run. 🪨✅ guideStone describes what the output means — that the computation’s results are the proof of their own correctness.

The Five Properties

A 🪨✅ guideStone-certified artifact satisfies five properties. Each is necessary. Together they are sufficient.

1. Deterministic Output — Same input, same binary, any hardware → same output within named tolerances. The binary detects its substrate (CPU, GPU, memory) at runtime and adapts its execution path, but the mathematical result is invariant.

2. Reference-Traceable — Every numeric claim traces to a source: a published paper, a mathematical proof, a physical constant, or a standard (ILDG, MILC, FAO-56, DADA2). No numbers float in space.

3. Self-Verifying — The artifact carries its own integrity. CHECKSUMS validated before execution. CRC on data payloads. Merkle roots when the provenance trio is wired. The consumer does not need to trust the transfer channel.

4. Environment-Agnostic — No hardcoded paths. No “install X first.” No platform assumptions. Pure Rust, static musl, cross-arch. Runs wherever the binary runs.

5. Tolerance-Documented — No magic numbers. Every threshold, every epsilon, every convergence criterion has a derivation. The derivation lives in the output metadata, not in a comment buried in the source.

The First Deployment Artifact

hotSpring-guideStone-v0.7.0 is the first artifact that left the development environment as a self-contained, self-verifying, self-benchmarking object.

What Was Built

A self-contained directory with:

• Dual-architecture binaries: x86_64 + aarch64
• Unified entry point: ./hotspring validate
• OCI container image for Windows/macOS/non-Linux
• Windows launcher: hotspring.bat (WSL2 → Docker fallback)
• Integrity manifest: CHECKSUMS (SHA-256)
• Self-knowledge: liveSpore.json (tracks every machine visited)
• Reference data + documentation

Cross-Substrate Validation

Cross-substrate observable comparison: 40/40 bit-identical. The physics does not depend on the instruction set, the C library, the GPU vendor, or the operating system.

Papers Validated

Three of TC Chuna’s published papers reproduced in pure Rust:

guideStone as Self-Leveling Benchmark

The deployment artifact has an emergent property: the validation is also a benchmark.

When ./hotspring validate runs on unknown hardware, it discovers the CPU architecture, probes GPU adapters, runs 59 physics checks, and reports wall time, throughput, and GPU utilization. The physics is the benchmark. The benchmark is the physics.

This means the artifact answers two questions simultaneously:

A PI evaluating a new GPU plugs in the USB, runs one command, and gets both answers. No installation, no configuration, no MILC build, no CUDA SDK.

The Onboarding Pattern

Traditional onboarding into computational physics workflows takes days to weeks: install toolchain, debug build failures, share input configurations, compare results via email, debug discrepancies, repeat.

🪨✅ guideStone onboarding takes minutes:

1. Receive USB or tarball
2. ./hotspring validate — 59/59 pass
3. ./hotspring benchmark — characterize hardware
4. ./hotspring generate --beta=6.0 --lattice=8x8x8x16 — produce data

The artifact is the conversation starter. The physics speaks for itself.

The Orthogonal Dimension

The ecosystem has three independent classification axes:

Structure — what is it built from? 1️⃣📦 UniBin → ♻️📦 ecoBin → 🧬📦 genomeBin

Deployment — where does it run? ⚛️🧬 NUCLEUS → 🌿📋 Niche → 🐭📡 fieldMouse

Verification — what does its output mean? 🪨✅ guideStone

These axes are independent. A 🐭📡 fieldMouse sensor node can be a 🪨✅ guideStone (its soil pH readings are calibrated against NIST standards). A full ⚛️🧬 NUCLEUS can produce output that is not 🪨✅ guideStone (exploratory computation with heuristic thresholds).

The Metrological Analogy

Before 2019, the kilogram was defined by a physical artifact — Le Grand K, a platinum-iridium cylinder in a vault in Paris. If Le Grand K changed, every kilogram on Earth drifted with it.

In 2019, the kilogram was redefined in terms of the Planck constant — a mathematical relationship that does not change, does not drift, and does not depend on any physical object. Any laboratory can realize the kilogram independently. The measurement became self-proving.

🪨✅ guideStone is the 🔧🦎 ecoPrimals equivalent. The consumer trusts the output because it is reproducible from the mathematics. The binary is the instrument. The output is the measurement. The mathematics is the invariant substrate.

Get the Artifact

The first 🪨✅ guideStone artifact — hotSpring-guideStone-v0.7.0 — is available for collaborators and anyone who wants to verify.

For collaborators and reviewers:

Contact ecoPrimal for a tarball, USB drive, or OCI container image. The artifact runs on any x86_64 or aarch64 Linux with zero dependencies. CPU-only validation takes approximately 3 minutes. GPU acceleration is automatic when a Vulkan-capable adapter is present.

For the curious:

• The spring repositories at syntheticChemistry contain the source — all public, all AGPL-3.0-or-later.
• 🧬📦 plasmidBin distributes pre-built primal binaries.
• The Science section has the 28 ⛺📄 baseCamp papers organized by domain.

The artifact is the conversation starter. The physics speaks for itself.

The wolf was always falling. The plaquette was always 0.59. The 🪨✅ guideStone does not create the truth. It makes the truth portable.