open-standards/mesh
Repository files (latest commit first)
Filename Latest commit message Latest commit date
Transaction Science 3e1c6d4465
All checks were successful
standards workspaces build + test / which workspaces this change touches (pull_request) Successful in 12s
standards workspaces build + test / macOS energy backends (pull_request) Has been skipped
repository checks / lockfiles, tests in CI, workflow hygiene, spec and section references, public text (pull_request) Successful in 1m42s
standards workspaces build + test / joulehook-rs (joulehook/joulehook-rs, joulehook-rs, linux) (pull_request) Successful in 1m13s
standards workspaces build + test / mesh-rs (mesh/mesh-rs, -p mesh-engine-gguf -p mesh-openai --features mesh-engine-gguf/gguf,mesh-openai/server, mesh-rs, linux-arm64) (pull_request) Successful in 2m23s
mesh: say what the energy figure covers — the served span, not the request
`Mesh::serve` routes at step 1, authorises at step 2, and opens the meter
session at step 3, immediately before `engine.serve`. So a receipt's energy is
the served span and excludes routing and authorisation. Nothing said so:
`grep -niE "routing (is|energy)|excludes? rout|before routing"` over
`mesh/SPEC.md` and `mesh-core` found nothing, and the field was documented only
as "the honest, provenance-tagged energy span" — which span was never stated.

That omission matters in the direction this family has already decided is
dangerous. The grant is debited **this** figure and admits further work on the
strength of it, so a shortfall under-charges the ceiling —
**under-reporting is the unsafe direction at a ceiling**, the rule written into
`jouleclaw/SPEC.md` §1.3 and propagated across the meters in #354. For a
table-driven router the gap is negligible. For a router that itself runs a model
— a classifier or an LLM choosing the route — it is not, and a reader taking a
mesh receipt's energy as the energy of handling the request would be wrong by
exactly that inference.

SPEC §3.3 now states the boundary, forbids presenting the figure as the
request's energy, and says that an implementation metering a model-running
router should seal that span as its own field so a verifier can still separate
the two. `MeshReceipt::energy` and the step-3 comment in `mesh-core` say the
same thing where someone reading the code will meet it.

**This states the limit; it does not remove it.** Metering the routing span
means moving the session boundary, which changes every receipt's figure and has
conformance consequences — a scope decision, not a doc fix, and the digest item
(sweep 2026-10-19 #17) carries it as a wedge rather than a defect.

**No test pins the ordering, and the spec says so.** The meter is
snapshot-driven and routing takes no samples, so the boundary is an ordering
invariant in the serving path rather than an observable property of a receipt.
Putting the router behind a trait, as the engine already is, is what would let a
test pin it; `Mesh` holds `router: Router` concretely today while `meter` is
already `&dyn Meter`.

One citation trap avoided while writing this: `mesh/SPEC.md` has its own §5 and
no §1.3, so a bare "JCP SPEC §5" would read as mesh's §5 and "jouleclaw SPEC
§1.3" would dangle. `check-spec-refs` scans only `.rs` files, so neither would
have been caught. Both cross-standard references name their file.

Verified: `cargo check --workspace --all-targets --locked` 0 errors, 22 test
binaries green with 0 failures, clippy over `mesh-core`/`mesh-receipt` exits 0
with no finding in a mesh file, `cargo doc` clean, `check-spec-refs` 1693
citations across 16 standards all resolving.

Co-Authored-By: Claude Opus 5 <[email protected]>
2026-10-21 05:42:26 -04:00
..
conformance/v1 mesh/1 + AR-1: receipts commit to the output; deterministic serves replay 2026-10-21 04:35:32 -04:00
mesh-rs mesh: say what the energy figure covers — the served span, not the request 2026-10-21 05:42:26 -04:00
tools/joules-per-decision mesh: withdraw the GPU per-decision figure — another process held 99.5% of the GPU 2026-10-05 07:20:58 -04:00
CONTRIBUTING.md wai.generate.spectral: extend the lexicon to 370 words (data-only, pin unchanged) 2026-06-27 05:33:47 -04:00
LICENSE wai.generate.spectral: extend the lexicon to 370 words (data-only, pin unchanged) 2026-06-27 05:33:47 -04:00
LICENSE-SPEC wai.generate.spectral: extend the lexicon to 370 words (data-only, pin unchanged) 2026-06-27 05:33:47 -04:00
README.md mesh: a grant is spent only by its issuer, at the steward it names, and its spend survives a restart 2026-10-17 17:13:42 -04:00
SPEC.md mesh: say what the energy figure covers — the served span, not the request 2026-10-21 05:42:26 -04:00

Mesh

Energy-accounted inference serving & routing. One of the Transaction Science open standards.

Mesh treats every inference as a metered transaction. A request declares a joule ceiling and a fidelity floor; the mesh routes it to the cheapest-sufficient model on the cheapest-sufficient backend, serves it under an energy meter, and seals a receipt whose headline figure is microjoules per accepted output token — measured at the meter, not estimated from a price list.

The serving shelf reports dollars per token and tokens per second. Mesh reports the number that is physical, comparable across hardware, and impossible to inflate: the joule.

The shape

ServeRequest ──▶ route (cheapest-sufficient) ──▶ authorise (JCP grant)
                                                     │
                                                     ▼
                     meter + debit ◀── serve (engine) ──▶ tokens
                                                     │
                                                     ▼
                         sealed JCR-1 receipt: µJ / accepted token

Four roles — caller, router, engine, steward — and four phases per serve: route, authorise, meter, seal. The grant is decided by JCP's own runtime before any work runs, so its expiry, caveats and remaining budget bind here as at any JCP enforcement point, and the metered joules are debited against it afterwards. A grant is spent only by its issuer, who signs each presentation for one request, and only at the steward it was issued to; each decision is journalled before the work, so a restart gives no budget back. See SPEC.md §2.

The reference workspace

mesh-rs/ is a pure-Rust reference implementation:

Crate What it is
mesh-schema wire types — ServeRequest, Fidelity, two-part Energy, RouteDecision, ServeOutcome
mesh-router energy-weighted cheapest-sufficient routing with a full candidate audit
mesh-engine the ServingEngine interface + a deterministic, weights-free ReferenceEngine
mesh-engine-gguf a real backend (behind gguf): wraps the family's pure-Rust GGUF/Metal inference substrate so the mesh routes and meters actual model tokens
mesh-meter the meter window; an always-available estimator + an optional bridge to the family hardware meter (RAPL / NVML / Apple) behind hw-meter, and ProcessMeter: this process's own CPU and GPU energy for an engine that runs in process, measured when a span covers the whole request
mesh-receipt the mesh/1 receipt, JCR-1 sealing (det-CBOR + COSE_Sign1), and a hash-chained log
mesh-core the orchestration that ties route → authorise → meter → seal together
mesh-conformance the deterministic reference scenario, its checks, and the gen-vectors bin

Mesh rides on JCP (a serve request is authorised by a JCP grant; receipts seal in JCR-1), dispatches to runtimes such as JouleClaw, and routes across compute described by EOC and isolated by Sandbox. It introduces no new cryptography (see SPEC.md §9).

Run it

cd mesh-rs

# all crates green
cargo test --workspace
cargo clippy --workspace --all-targets

# end-to-end: mint a JCP grant, serve, print the sealed receipt
cargo run -p mesh-core --example demo

# on Linux/RAPL or a CUDA/NVML box, meter real silicon (HwShunt provenance, where
# the counter is live; a counter that is not live falls back to the estimator)
cargo run -p mesh-core --example demo --features hw-meter-rapl

# regenerate the published conformance vectors
cargo run -p mesh-conformance --bin gen-vectors -- ../conformance/v1

# serve a REAL model end-to-end through Mesh (Falcon-H1 0.5B), seal the receipt
cargo run --release -p mesh-engine-gguf --features gguf --bin serve_falcon -- \
    /path/to/falcon-h1-0.5b-instruct 12 "The capital of France is"

The serve_falcon run demonstrates the whole standard on real model tokens: a JCP grant authorises each request (an underfunded grant is refused before any decode runs), the cheapest-sufficient backend is chosen, the decode is metered, and a mesh/1 receipt is sealed and hash-chained across the session — e.g. prompt "The capital of France is" → " Paris. It is known for its historical landmarks", prompt "Photosynthesis is the process by which" → " plants convert sunlight into chemical energy in the form of", with both seals verifying and the chain's prev-links intact.

The demo prints the route (every backend weighed, with the reason each was kept or rejected), the honest energy provenance, and the wedge metric — then verifies the sealed envelope under the steward key.

Status

v0.1.0-draft.1. Reference workspace builds and tests green (mesh-schema, mesh-router, mesh-engine, mesh-meter, mesh-receipt, mesh-core, mesh-conformance); cargo clippy --all-targets warning-free; published vectors in conformance/v1/. The reference engine is deterministic and weights-free; production backends implement the same ServingEngine interface — mesh-engine-gguf (behind gguf) is a real one over the family's pure-Rust GGUF/Metal substrate, kept out of the default build so it needs no weights.