The IOReport liveness rule said Live as soon as ONE CPU rail sustained its
advances (`Quorum::Any`), but `snapshot` sums every CPU rail into the one
`Cpu` figure. On a multi-die part that is two rails, `DIE_0_CPU Energy` and
`DIE_1_CPU Energy` on an M3 Ultra, so one die live beside a frozen one would
have selected the probe and sealed one die's energy as the whole CPU's, as
`HwShunt`. RAPL and NVML were already held to every rail they sum
(6b693aa4); IOReport now is too.
The rule: Live only when every CPU rail present in the samples advanced in
at least five intervals, four in five of those observed, over at least
300 ms; Frozen otherwise; Absent with no CPU rail. A rail is present if it is
in any sample the probe read, the baseline included, and it is judged over
every interval observed. Before, a rail seen only in the first sample was
never counted, because rails were registered from the later sample of each
interval; now the baseline's rails are seen with no advance. That holds for
the RAPL and NVML runs too, which share the watch. With one rule left there
is no quorum to choose, so `Quorum` and `watch_with` are gone: `RailWatch` is
the rule, and `counter_run` is `watch` over the counter.
The oracle behind the real-IOReport tests now says "moved" only when every
CPU rail it reads changed at four reads in five or more, so a Live claim is
checked against every die. It reads exactly the probe's rails on both hosts:
one on this M5 Max, the two dies on the M3 Ultra.
Docs: the liveness and ioreport module docs, `CpuRail`, `RailWatch`,
`IoReportStatus` and its Display text, `ioreport_status`, the probe module
doc, and the sandbox README's metering guarantee.
Tests, all scripted (macOS here). The rule:
`two_dies_that_both_keep_advancing_are_live` (live at the tenth interval, or
at the fifteenth when DIE_1 misses its first three),
`one_die_that_keeps_advancing_beside_a_frozen_one_is_frozen` (the other die
held, stepped once, stepped five times, or advanced three intervals in five;
in either listing order; and twin `CPU Energy` channels),
`one_die_that_keeps_advancing_with_no_other_die_in_the_samples_is_live`, and
`a_die_in_only_some_samples_is_a_rail_the_watch_needs`. A whole probe run:
`a_probe_run_calls_two_dies_that_keep_advancing_live`,
`a_probe_run_calls_one_live_die_beside_a_frozen_one_frozen`,
`a_probe_run_calls_one_die_alone_that_keeps_advancing_live`. The production
status through the scripted IOReport seam:
`the_production_status_of_two_dies_that_keep_advancing_is_live`,
`the_production_status_of_one_live_die_beside_a_frozen_one_is_frozen`. The
single-rail tests are unchanged. `a_counter_with_one_rail_that_held_is_frozen`
asserted that the IOReport loop called its two-rail samples live; it now
asserts Frozen.
sandbox-meter: 106 lib tests and both once-per-process files pass on this M5
Max. The same lib test binary passes 106/106 on the M3 Ultra (macOS
27.0.1), where the real probe gave `CpuRailFrozen` with `DIE_0_CPU Energy`
and `DIE_1_CPU Energy` at 0 advances each in 28 intervals over 991 ms, and
the oracle saw 0 changes on both dies in 17 reads. The sandbox-rs crates that
depend on the meter pass (126 tests), as do jcp-gateway and
jcp-action-receipt (33), joule-code-remediate (13), and joule-ui-interaction
and joule-ui-serve (19). joule-ui `gen_vectors` regenerated into a scratch
directory is byte-identical to `conformance/v1/`. Clippy is clean for
aarch64-apple-darwin and x86_64-unknown-linux-gnu.
Falsified, each restored from git and the suite re-run green: the rule back
to any one rail fails the three one-live-die-beside-a-frozen-one tests (rule,
run, production status), `two_dies_that_both_keep_advancing_are_live`,
`a_die_in_only_some_samples_is_a_rail_the_watch_needs` and
`a_counter_with_one_rail_that_held_is_frozen`; not seeing the baseline's
rails fails `a_die_in_only_some_samples_is_a_rail_the_watch_needs`.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
13 KiB
Sandbox
A sandbox for the heterogeneous compute fabric. Isolate, capability-gate, meter, and attest — owned by no one.
Sandbox is the isolated-execution substrate of the Transaction Science open standards. Where Smart Byte carries value, EOC carries compute, and Joule Code carries coding work, Sandbox carries the place those operations run: a bounded environment in which an untrusted unit of work executes, sees only what it was granted, and leaves behind a signed account of what it did and what it cost.
Every other sandbox in the field isolates a CPU process — a container, a microVM, a Linux process pinned to a core. The accelerator, where present, is an opaque passthrough that nobody meters and nobody attests. But agentic and AI work does not run on a CPU. It runs across a fabric — CPU, GPU, NPU, TPU, photonic, quantum. So in this standard the unit of isolation is not a process; it is a heterogeneous unit of work, and the accelerator is a first-class citizen of the isolation boundary: a capability you grant, a resource you meter in joules, and a line item you attest.
A Sandbox run carries exactly four guarantees: isolation (the unit runs at the lowest tier of an eight-rung ladder, S0–S7, that satisfies its needs, and cannot reach past it), capability (every resource the unit touches — each device class included — was explicitly granted; ungranted access is denied), metering (the run emits a signed energy receipt with per-device-class joule attribution and an honest provenance tag on every reading — and the reference meter reads vendor energy counters: Apple Silicon's energy model via IOReport, Intel/AMD RAPL via powercap, NVIDIA's total-energy counter via NVML. A reading is tagged HwShunt only while its counter is demonstrably live: each vendor counter is used only if every rail of it that the meter sums into one figure keeps advancing while the meter's probe burns CPU — in at least 5 sample intervals, at least 4 of every 5, over at least 300 ms — every IOReport CPU energy rail (the rail itself, one per die on a multi-die part), every RAPL package and every NVML GPU; and a span across which the CPU counter held is never a reading. A counter that is absent, or present but not live — frozen, or republished in lumps, as the IOReport CPU rails are on macOS 27 — yields no hardware-graded figure: the span degrades to the labeled Estimator (wall-clock × TDP), never a fabricated measurement, and ioreport_status(), rapl_status() and nvml_status() say why a host is estimated. Because a live counter reports the whole SoC, the meter can attribute it to a single process tree by CPU-time apportionment — tagged ModelBased, since the reading is a measurement but the split is a model), and attestation (a signed statement of what image ran, at what tier, under what grants, on what hardware, for what cost, with a deterministic-replay journal hash where one was produced).
The substrate is intentionally ownerless: no protocol-level token, permissive licensing on spec text and reference code, and conformance vectors so any implementer can verify canonicity without a trust relationship with the authors. A conforming host runs a unit, enforces its grants, meters its energy, and seals its attestation without contacting any single steward's servers as a condition of operation. Transaction Science is one steward: it publishes the spec, ships a reference implementation, and operates commercial services around the substrate — not its proprietor.
The doctrine
Isolation per joule, across the whole fabric. The durable layer of a sandbox is not fast container boot or a managed control plane — both are commoditized. It is the boundary itself, made honest: the lowest isolation tier that satisfies the work, every resource gated by an explicit capability, and a signed receipt that attributes energy to each device class it touched. "Run it safely" and "don't waste energy" are the same optimization: an over-isolated run burns joules it did not need, and an unmetered accelerator is exactly where the energy hides.
Three consequences, normative in SPEC.md:
- The ladder. A host MUST run a unit at the lowest isolation tier of
S0:Bare → S1:Capability → S2:Sandbox → S3:Partition → S4:Unikernel → S5:MicroVM → S6:Confidential → S7:Sovereignthat satisfies the unit's capability set, and MUST NOT escalate beyond what the grants require. Escalation is the escape hatch, not the entry point — the same discipline the cascade applies to inference, applied to isolation. - The grant. Every resource a unit can touch — network, storage, key-value, and each compute device class (
Cpu | Simd | Gpu | Npu | Tpu | Lpu | Photonic | Quantum) — MUST be explicitly granted. Access to an ungranted resource MUST be denied. Each compute grant carries a joule budget. - The attested receipt. Every completed run MUST emit an energy receipt accounting joules per device class —
measured + estimated, integer microjoules — each reading taggedHwShunt | ModelBased | Estimator, sealed as a JCR-1 envelope; and MUST emit an attestation binding the unit, image, tier, grants, hardware, and receipt, signed Ed25519 over JCS-canonicalized JSON — the same attestation primitive the ARL Sandbox (ARL-S) uses, so the two interoperate at the attestation layer.
Contents
SPEC.md— the normative specification (RFC 2119): the unit of work, the S0–S7 isolation ladder, the capability-grant model, the per-device energy receipt, the attestation, and the conformance contract.spec/— sub-specifications:tiers.md(the ladder),capabilities.md(the grant model),receipts.md(per-device energy),attestation.md(the signed claim + replay journal),conformance.md(test obligations).conformance/— signed wire vectors. An implementation is conformant if it round-trips these.proof/— the proof matrix: a ledger of signed proof cells, each a real run of the conformance pack + a sealed workload on a real target (substrate × OS × arch × cloud × hardware × tier). A green cell is a signature over a real run, not a claim.sandbox-rs/— the Rust reference library: content addressing and tiers (sandbox-schema), the capability model and deny-by-default enforcer (sandbox-capability), the per-device receipt and JCR-1 sealing (sandbox-meter), the attestation and replay journal (sandbox-attest), the reference host — tier minimization plus sealed-run assembly (sandbox-runtime), a real S2 (WASM) executor on wasmtime with fuel→joule metering and a brokered, Enforcer-gated host-call surface that journals every nondeterministic input for replay (sandbox-exec-wasm), a real S5 (microVM) executor driving Firecracker — both an API-socket driver and a supervised-child lifecycle driver — with pluggable per-device power metering (sandbox-exec-microvm), the S6 confidential-tier TEE layer — SEV-SNP report model, policy + real ECDSA-P384/SHA-384 verification, and aconfigfs-tsmhardware provider with a software-grounded test provider (sandbox-tee), the conformance harness with thegen_vectorsgenerator (sandbox-conformance), the proof-matrix harness that runs the pack + a sealed run on a target and emits a signed proof cell (sandbox-proof), thesandboxreference checker/runner (sandbox-cli), and the JCP gate (sandbox-jcp) — execution as a JCP consumer: a metered Grant bound to the unit's image by content address is decided by the JCP runtime before the executor is ever invoked, the declared energy ceiling is debited from the delegation chain pre-run, and the decision (allow and deny alike) seals as a JCR-1 receipt alongside the sandbox's own sealed run.
Heterogeneity is normative, not an extension
The clause that distinguishes this standard from a process sandbox is in SPEC.md §2.3 and §6.2: where a unit is granted an accelerator capability, a conforming host MUST attribute energy to that device class separately in the receipt, MUST record the device identity (class, model, driver) in the attestation, and at tiers S5 and above MUST report the measurement provenance honestly (HwShunt where the device exposes a live energy interface — NVML, RAPL, ROCm SMI, Level Zero — ModelBased or Estimator otherwise). A host that runs accelerated work but reports a single undifferentiated CPU energy figure is not conformant.
Lineage
Sandbox draws its design from two Transaction Science runtimes and generalizes a third standard:
- The isolation ladder and the capability-grant model follow the layered, hardware-rooted isolation and the per-resource grant set proven in the invisible-infrastructure runtime — capabilities first, WASM and microVM tiers above, TEE and sovereign tiers at the top, accelerators as explicit device capabilities throughout.
- The per-device energy receipt and its provenance discipline follow the picojoule receipt and measurement-quality model of the JouleDB substrate, sealed in the family's shared JCR-1 envelope.
- The attestation generalizes the SUT / Harness / Supervisor separation and the signed-telemetry model of the ARL Sandbox out of a measurement-only role into a standalone substrate that ARL-S, JouleClaw, EOC, and Joule Code can target instead of each re-specifying isolation.
Status
v0.1.0 draft. Specification core published (the unit, the ladder, the grant model, the per-device receipt, the attestation); the reference library lands the schema, capability, meter, attest, runtime, and conformance crates with the sandbox CLI, plus two real tier executors: a wasmtime-backed S2 executor that runs untrusted WASM under fuel and memory caps, meters fuel→joules, and brokers every capability the guest exercises (kv, clock, randomness) through the deny-by-default Enforcer — journaling each nondeterministic input so the run's attestation carries a replay-journal root; and a Firecracker-driver S5 executor that boots a microVM (API-socket or supervised-child lifecycle) and meters CPU and an assigned accelerator separately through pluggable power meters (RAPL/NVML-class, with an honest ModelBased fallback). Energy receipts seal as JCR-1 envelopes (canonical det-CBOR + COSE_Sign1, EdDSA), the signed-receipt format shared across the family; attestations sign Ed25519 over JCS-canonicalized JSON, interoperable with ARL-S. The conformance vectors in conformance/v1/ are real, cryptographically computed values — including a committed pack of adversarial negatives a conformant host must reject — regenerated by gen_vectors and verified by the harness (in-code and file-backed).
Readiness is classified per component (Solid / Engineering / Research / Theory) in SPEC.md §9 and respected when sequencing work. The S2 (WASM) executor runs and meters real guests; the S4–S5 (microVM) driver is real Firecracker integration (it boots a guest on Linux+KVM); and the S6/S7 confidential and sovereign tiers are implemented end-to-end — the host obtains a TEE report, verifies it (real ECDSA-P384/SHA-384 signature + measurement/TCB/debug/report-data policy), seals the evidence into the attestation, binds a locality, and links the run into a tamper-evident audit chain. The single honest caveat (SPEC §9): generating a genuine vendor-signed report requires the CPU's security processor — the configfs-tsm provider does so on real confidential hardware and reports Unavailable elsewhere, while a tagged software-grounded provider, which never claims a hardware root, covers wiring and tests.
Related
- ARL — the AI Readiness Level standard; its ARL Sandbox (ARL-S) is the measurement-specialized profile of this substrate. The attestation primitive (Ed25519 over JCS) is shared, so an ARL-S session attestation and a Sandbox run attestation verify under the same tooling.
- JouleClaw — the energy-optimised AI runtime; a JouleClaw agent's tool execution runs inside a Sandbox, and the receipt provenance discipline (
HwShunt | ModelBased | Estimator) and the JCR-1 envelope are JouleClaw's. - Joule Code — a coding operation that executes generated code (a test run, a build) runs that code inside a Sandbox; the joule receipt composes.
- EOC — the energy-optimised compute substrate; the discipline of stopping at the cheapest tier that closes the request is EOC's, applied here to isolation tiers.
Licence
Reference code (everything under sandbox-rs/) is Apache-2.0 (LICENSE). Specification text (SPEC.md, spec/) is CC-BY-4.0 (LICENSE-SPEC). The wire format and the right to fork are public.