Evidence Snapshot

Canonical metrics for the ecoPrimals ecosystem — single source of truth for all counts, definitions, and measurement methodology. Every other page should agree with this one.

Purpose

This page defines every metric used across sporePrint. When a number appears elsewhere on the site, it should either pull from this registry via shortcodes or state the measurement date explicitly. If a page conflicts with these numbers, this page is correct and the other page is stale.

Measured: 2026-08-04-PM — via spore-validate refresh (tokei line counts + cargo test pass counts from source repos)

Ecosystem Scale

MetricValueDefinition
Total Rust LOC3,598,358Lines of Rust counted by tokei across all primal + spring repos
Primal Rust LOC2,719,240Infrastructure code ( 15 primals)
Spring Rust LOC879,118Science validation code ( 9 springs)
Total test functions135,000+cargo test unit + integration tests across all repos
Primal tests86,240Infrastructure test functions
Spring tests34,760Science validation test functions
WGSL shaders952 files, 74K linesVendor-agnostic GPU compute (WebGPU)
Validation checks20,695+Quantitative science assertions with explicit numerical tolerance
Papers reproduced175+External peer-reviewed publications whose results are reproduced in Rust
baseCamp papers28ecoPrimals’ own executable manuscripts/studies
Primals15Sovereign infrastructure binaries (Rust, statically linked)
Springs9Domain-specific science validation environments
Content pages338Pages on this site (sporePrint)

What These Numbers Mean

Test functions vs. validation checks

Test functions are standard Rust #[test] functions counted by cargo test. They include unit tests, integration tests, property tests, and fuzz harnesses. The number 135,000+ is the sum of all cargo test passes across all repos.

Validation checks are a subset: the 20,695+ quantitative science assertions that compare computed results against published values with explicit numerical tolerances. These are the “does the science reproduce?” checks. Every validation check is also a test function, but not every test function is a validation check.

Papers reproduced vs. baseCamp papers

Papers reproduced ( 175+) are external, peer-reviewed publications from journals (Nature, Science, PNAS, etc.) whose key results are reproduced in Rust with explicit tolerance comparisons. The count includes papers across all 9 springs.

baseCamp papers ( 28) are ecoPrimals’ own executable manuscripts — each is a narrative with embedded cargo test results that a reader can reproduce. These are in the Science section.

Primals vs. springs vs. products

Primals are infrastructure: the binaries that form the mesh ( Songbird for routing, BearDog for identity, NestGate for storage, etc.). There are 15.

Springs are science validation environments: domain-specific test suites that reproduce published results. There are 9 (7 science domains + neuromorphic hardware + 1 meta-spring for ecosystem validation).

Products are compositions of primals aimed at specific use cases ( helixVision, blueFish, esotericWebb). Products have their own maturity levels.

Measurement Methodology

All metrics come from source code, not estimates:

  1. LOC: tokei run on each repo’s src/ and crates/ directories
  2. Tests: cargo test pass counts from CI or local runs
  3. WGSL: tokei on *.wgsl files in barraCuda
  4. Validation checks: counted from validate_* and exp_* test binaries
  5. Papers: counted from spring validation summaries (each paper has a named test)

The spore-validate refresh command automates this: it clones all repos, runs tokei, and compares against the registry. Drift beyond 5% triggers a warning.

Maturity Levels

Claims across this site carry maturity labels:

Implemented — Code exists and tests pass.

Reproduced — Matches an external published result with explicit tolerance.

Certified — Portable guideStone artifact exists and is verifiable.

Architecture-ready — Design exists and is partially implemented; not fully validated.

Planned — Roadmap item. No implementation yet.

Unaudited — Claim is not externally reviewed (security, compliance, regulatory).

When you see a claim without a maturity badge, assume Implemented for code claims and Unaudited for compliance/security claims.

Safety Model

#![forbid(unsafe_code)] is enforced at the crate root of all spring crates, all provenance crates ( BearDog, NestGate, Squirrel), the validation binary (spore-validate), and most infrastructure crates.

The exception is ToadStool: GPU and NPU dispatch requires unsafe FFI at hardware boundaries. toadStool contains documented, safety-audited unsafe blocks confined to hardware-containment crates. These are explicitly scoped, individually justified, and isolated from the rest of the ecosystem.

Dependency chain: No C/C++/Fortran libraries appear in the runtime dependency chain. blake3 uses the pure feature (Rust-only); flate2 uses rust_backend. The cc crate appears only as an unused build dependency.

Historical Notes

Pages dated March 2026 reflect the ecosystem state at that time (~3.2M LOC, ~107K tests, 7 springs, 14 primals). Those pages are historical snapshots. The current numbers are on this page.

Pages that use {{ "{{" }} total_stat(...) {{ "}}" }} or {{ "{{" }} entity_metrics(...) {{ "}}" }} shortcodes pull from the live registry and are always current.

Per-Entity Metrics

For individual primal and spring metrics (LOC, tests, files, crates), see:

Verify It

git clone https://github.com/ecoPrimals/sporePrint.git && cd sporePrint
cargo run --manifest-path crates/spore-validate/Cargo.toml -- validate --check --verbose
cargo run --manifest-path crates/spore-validate/Cargo.toml -- certify

The certify command computes a BLAKE3 Merkle root of the entity graph. Compare it against the published manifest. If they match, the registry is internally consistent.