The monorepo's IOReport energy backends now decline unless the CPU energy rail keeps advancing: on macOS 27 an unprivileged reader sees the CPU rails frozen or republished only in rare lumps, so a meter that trusted a present counter reported CPU work as GPU-only energy, or as zero, while tagging it a reading. The owner's decision is that a vendor energy counter is a real reading (the trust gap is provenance, not accuracy), so a LIVE reading is graded as a hardware reading, and every spec that grades vendor counters now carries the same rule: A reading from a vendor energy counter (RAPL, NVML, Apple IOReport, ...) MAY be graded as a hardware reading only while the counter is demonstrably live for the domain the figure claims: the implementation MUST have observed the counter covering that domain advance under load (for Apple IOReport: the CPU energy rail itself advancing in at least 5 sample intervals, in at least 4 of every 5, over at least 300 ms while the probing thread burns CPU). A counter that is present but not live - frozen, or republished in lumps - MUST NOT produce a hardware-graded figure; the implementation falls back to its estimator or reports the work unmetered, and SHOULD record why. A figure MUST NOT claim coverage of a domain its counter does not read; where the format has a coverage field, such a figure is Partial. Joule Code - spec/receipt.md §2 (REQUIRED): HwShunt now reads "an external shunt, or a vendor energy counter (RAPL, NVML, Apple SMC/IOReport) while that counter is live"; the rule above is added verbatim, and in this receipt the hardware grade is a present measured_uj and the HwShunt tag, so a counter that is not live supplies neither. The §1 comment and §5 say "live". - SPEC.md §6: the provenance definition now names IOReport and NVML beside RAPL/SMC, measured_uj comes only from an interface live for the domain it claims, and a paragraph states the MAY/MUST NOT, pointing to receipt.md §2 for the test. The §9 readiness row says "live" and "not live". - ROADMAP.md v0.2 "Measured energy": a vendor counter feeds the HwShunt path only while live; where it is present but not live (IOReport's CPU rails to an unprivileged reader on macOS 27) the receipt stays on the estimator. - examples/l3-jouleclaw/README.md: the claim that the receipt is "a real local measurement ... (HwShunt on RAPL/NVML hosts; Estimator on Apple Silicon until IOReport is wired)" was wrong both ways: run.rs wires only Apple Silicon's counter, and other hosts fall to the token Estimator. It now says so, that the counter's own tag is carried, and that HwShunt needs a live counter. - joule-code-receipt: the Provenance::HwShunt doc comment says the same. - CHANGELOG.md [Unreleased] ### Changed: the rule; no wire or vector change. Sense - SPEC.md §7 (REQUIRED): measured_uj comes only from a live interface; the rule is added verbatim. The Sandbox PowerMeter was described as "(real RAPL/IOReport)"; it is now described as grading a reading HwShunt only while its counter is live and falling back to its estimator otherwise, which is what origin/sandbox-meter/cpu-rail-liveness implements. §10's Sandbox edge says the same in brief. - sense-schema: the HwShunt, Energy and Energy::measured doc comments say a vendor counter is a hardware reading only while live. - Sense keeps no CHANGELOG, so none is added. Other specs that grade vendor counters - map/spec/receipt.md §2: the HwShunt row read "RAPL/SMC/shunt"; it now names a counter "while it is live", and the rule is added verbatim, noting that Receipt::seal cannot check liveness (the meter's duty, upstream). map/SPEC.md §6 states the MUST NOT and points there. CHANGELOG.md gains an [Unreleased] ### Changed entry; no wire or vector change. - arl/ARL-S.md, Telemetry: physical telemetry is what gates tiers above 0, and it said energy is "measured directly via RAPL ... NVML". The rule is added in ARL-S's own terms and lower-case keywords (ARL uses no RFC 2119 capitals): a counter that is not live does not produce physical telemetry, the energy it would have covered is reported unmetered, and the Supervisor must have observed the counter advance. No version bump. - smart-byte treatise §8.6: "Where RAPL or equivalent hardware energy counters are available, the measurement is read from the hardware" now requires the counter to be demonstrably live, and a frozen or lump- republished counter yields the same honest zero as an absent one. Text and doc comments only: no code path, wire format, or conformance vector changes. joule-code-rs (joule-code-receipt, joule-code-conformance) and sense-rs (sense-schema, sense-conformance) build and test clean, their rustdoc builds under -D warnings, map-rs (map-receipt, map-conformance) tests pass, `joule-code-conformance verify ../conformance/v1` passes 24/0/0, the no-vanity-metrics check passes, and every relative link in the changed Markdown resolves. Co-Authored-By: Claude Opus 5.5 <[email protected]>
14 KiB
Sense v0.1 — A Standard for Ambient Context
| Standard | sense |
| Status | v0.1.0 draft for review |
| Licence | Specification text CC-BY-4.0 (LICENSE-SPEC); reference code Apache-2.0 (LICENSE) |
| Reference | sense-rs/ |
A vendor-neutral standard for the ambient context substrate — the always-on, deterministic, energy-accounted record of what is happening around an agent, and a principled account of when that is worth a human's or another agent's attention. Where the tool protocols (MCP, A2A, ACP) standardize an agent reaching for capabilities, Sense standardizes the afferent half: the stream of observations flowing in, and the governed decision to surface them.
1. Scope and doctrine
- Afferent, not efferent. Sense records the world the agent is in — file changes, focus, logs, thermals, shell activity — not the agent's own reasoning trace. The reasoning trace is observability's domain (OpenTelemetry GenAI), is non-deterministic by its own admission, and is out of scope here.
- Deterministic by construction. Every observation is content-addressed and totally ordered under a hybrid logical clock. Replaying a log yields a byte-identical projection on any machine (§3). Determinism is what makes an ambient record auditable rather than merely logged.
- Perception ⊗ attention-budget. Observing is cheap; interrupting a human is not. Sense pairs a salience projection (§4) with an attention budget (§5) that decides whether an observation surfaces, scoring its value against a modelled interruption cost and pricing the surfacing in joules. This is the afferent dual of JCP's capability ⊗ joule-budget.
- Inference is the last resort. A question about the observed world is answered from the log by a deterministic recall cascade; a model is called only when nothing cheaper can answer (§6). Cheapest sufficient answer wins.
- Provenance and trust are declared, never promoted. Every datum carries its source class and a trust level (§2); every energy figure carries its provenance (§7). A host may downgrade what it is unsure of; it must never silently upgrade.
- Non-goals. Sense does not mandate a sensor set, a UI, a model, a transport, or a storage engine. It standardizes the observed datum, the determinism, the attention decision, the cascade accounting, and the receipt — not the daemon.
2. The observed event (REQUIRED)
An Event is one datum on the log:
hlc— a hybrid logical clock stamp; the event's place in the total order.source— a SourceClass: the vendor-neutral class of the signal (file_change,app_focus,system_log,process_tree,thermal_power,shell,accessibility,network,endpoint,salience). A host maps its native sensors onto these classes — macOS FSEvents, Linux inotify and WindowsReadDirectoryChangesWall surface asfile_change— so a log means the same thing on every platform. Codes are stable and append-only; vendor-specific classes usevendor:<code>withcode ≥ 1024.valid_time_ns— optional wall-clock time at observation, captured as data, not as the clock (so replay stays deterministic when the wall clock jumps).trust— a TrustLevel:Attested(cryptographically vouched, e.g. a Sandbox attested capture),FirstParty(a first-party OS API the host called),Inferred(parsed from an untrusted byte stream), orUntrusted(the default).payload— the content address (§3) of the raw bytes; the bytes live in the store's content-addressed table, recoverable from the address alone.line— the textual projection the attention and synthesis tiers read.
An event's content id is blake3 over a fixed-order, length-prefixed
concatenation of its fields. The id binds the source class and trust level, so an
event cannot be re-classified or re-trusted without changing its id (conformance
C2).
3. Canonical serialization and determinism (REQUIRED)
- Content address. A
Cidisblake3(payload_bytes), lowercase-hex on the wire. Identical bytes always yield the identical address; storage is the index. - Total order. Events are keyed by an
Hlc = (physical_ns, logical). The physical component is a monotonic anchor;logicalbreaks ties at the same instant. The clock is totally ordered within a daemon run; on restart a host seeds it past the maximum stored stamp so new stamps strictly exceed all stored ones. - Replay determinism.
events_since(Hlc::ZERO)replays the whole log in clock order. Two hosts fed the same event sequence produce a byte-identical replay digest —blake3over each event's content id in order (conformance C1). A production host MAY persist to an MVCC B-tree keyed by the HLC big-endian bytes; the wire and the order are fixed here, not the engine. - Receipt core (§8) uses canonical det-CBOR (RFC 8949 §4.2.1: length-first then bytewise key ordering, shortest integer form, no floats/bools/nulls), the same encoding the family's JCR-1 receipt signs over.
4. The salience projection (REQUIRED)
A bounded-state observer consumes the event stream and projects three scalars
in [0, 1]:
- salience — is the current source bursting? (the fraction of a recency window occupied by the current source class).
- novelty — is the current source rare? (one minus its share of all traffic seen).
- uncertainty — is the recent window low-trust? (the fraction of the window
at
Inferred/Untrusted). An honesty signal: "the system isn't sure what's going on."
The reference observer computes these from exact integer counts, so two runs over the same events project byte-identically. A host MAY substitute a model observer (e.g. a recurrent bounded-state network); the standard requires only that a conforming observer be deterministic under replay, and that the three scalars carry these meanings.
5. The attention budget (REQUIRED)
A continuous projection is not yet a decision. A detector turns it into discrete Highlights by an EMA-baselined k-sigma rule with an absolute floor and a debounce — it fires on anomalies relative to a probe's own recent baseline, above an absolute floor, not on every reading. The decision is made against the prior baseline (then the baseline updates) so a single large spike does not suppress its own fire.
A highlight is then gated by the attention budget before it reaches a human or agent. Following the proactive-interruption literature, the gate scores:
- value — the highlight's probe score, with
uncertaintyhighlights discounted (an honesty signal is usually less worth an interruption than a true salience or novelty spike). - interruption cost — displacement (proportional to how engaged the user is) plus interference (a penalty when the surfacing would occupy the user's currently-active modality — a visual alert while the user reads).
A highlight surfaces only when value > cost + margin and the energy
budget can pay the surfacing's joule price. The three outcomes are surface,
hold:value (value did not clear cost), and hold:budget (no joules left). The
value test comes first: a highlight that is not worth an interruption is
hold:value whether or not the budget could pay for it, so hold:budget means a
worthy highlight the budget could not afford. The decision is deterministic in its
inputs (conformance C3). Surfacing debits the
budget; a hold spends nothing. This makes attention a metered resource: an
ambient agent cannot interrupt without account, and cannot interrupt past its
energy budget at all.
6. The synthesis cascade (REQUIRED)
When a human or agent asks a question of the log, the host builds an Envelope — the top-k retrieved spans + the salience projection + the query — and walks a speculative cascade:
- Deterministic recall first. A Tier-0 stage recognises a small set of factual-recall intents (recent files, last command, what was surfaced, …) and answers directly from the envelope's spans, with no model call. It is conservative: it answers only when query-side and span-side evidence agree, and escalates when unsure (a missed answer costs a paid call; a wrong answer is worse).
- Escalate as a last resort. Only when no cheap stage answers does the request reach the fallback — typically a frontier model.
Retrieval is BM25-lite over exact integer term frequencies, then
source-stratified recency fill for any empty slots (natural-language queries
rarely overlap span text by token, and an empty envelope answers nothing). Spans
are content-addressed, so an envelope built from the same log + query is
byte-identical and a provider-side prompt cache hits cleanly. Every answer carries
an honest Energy figure (a deterministic answer is an
honest zero estimate, never a measurement).
7. Energy and provenance (REQUIRED)
Every metered figure — a surfacing's price, a synthesis's cost — is an integer
microjoule pair: measured_uj (present only from a live hardware interface) and
estimated_uj (always present), tagged with a Provenance (HwShunt,
ModelBased, Estimator). This is the family's shared energy model: a figure is
reported with the honesty of how it was obtained and never promoted. The
honesty invariant is that HwShunt requires a measurement and Estimator
forbids one.
A reading from a vendor energy counter (RAPL, NVML, Apple IOReport, ...) MAY
be graded as a hardware reading (HwShunt, with measured_uj) only while the
counter is demonstrably live for the domain the figure claims: the implementation
MUST have observed the counter covering that domain advance under load (for
Apple IOReport: the CPU energy rail itself advancing in at least 5 sample
intervals, in at least 4 of every 5, over at least 300 ms while the probing
thread burns CPU). A counter that is present but not live — frozen, or
republished in lumps — MUST NOT produce a hardware-graded figure; the
implementation falls back to its estimator or reports the work unmetered, and
SHOULD record why. A figure MUST NOT claim coverage of a domain its
counter does not read; where the format has a coverage field, such a figure is
Partial.
Measurement itself is a recognized edge: the sibling Sandbox standard's
PowerMeter is the reference meter — it reads RAPL, NVML or IOReport, grades a
reading HwShunt only while its counter is live, and falls back to its estimator
when it is not. Sense defines what an interaction records; the host wires the
meter it trusts.
8. The sealed, hash-chained surfacing log (REQUIRED)
An ambient agent's audit duty is the inverse of a tool agent's: it must show it did
not observe-and-surface without account, and that no record of what it surfaced
(or chose to hold) was quietly removed. Every gate decision (§5) therefore seals
as the next link of a hash chain. The signed core commits to the source
event's content id, its trust level, the highlight, the value/cost the budget
computed, the decision, the joules charged, and the budget left; each link's
prev is the BLAKE3 content id of the previous link's canonical core, and seq
increments from zero.
Allows and holds are both chained — a held highlight is exactly the record an
adversarial host would prefer to forget. Excising, reordering, or editing any link
breaks the chain at that point (conformance C5). The core signs with the host's
Ed25519 key; verify_chain checks every signature, the sequence, and every prev
link, and reports the first break. A link MAY additionally be expressed as a JCR-1
COSE_Sign1 envelope for verification by the family's shared cross-language tooling
(a recognized edge, not required — the hash chain is the primary audit).
9. Conformance
A conforming implementation reproduces, from the published vectors alone
(conformance/v1/):
- C1 — capture determinism. A fixed event sequence replays to the published
replay_digest; two builds agree. - C2 — provenance & trust. A fixed event's content id matches the published
event_id; re-tagging its source class or trust level changes the id. - C3 — attention. A fixed signal stream yields the published sequence of
highlights and surface/hold decisions, and the same highlights gated for a busy
user on a small budget yield every outcome —
surface,hold:value,hold:budget— with the published budget after each. - C4 — synthesis cascade. A recall query is answered with no model call
(
deterministic); an open-ended query escalates. - C5 — chain integrity. A surfacing log verifies (
verify_chain → n), and excising the middle link breaks it at the published index.
10. Recognized edges
Sense is one organ in the family; each boundary it leans on is a published sibling standard, not a dependency it owns:
- JCP (capability ⊗ joule-budget) — gates who may subscribe to which source class, for how many joules. Sense is the afferent dual.
- Sandbox — runs the capture in metered, attested isolation; its
PowerMeteris the reference meter (§7: a vendor counter only while live, else its estimator) and its attestation is what earns a datum theAttestedtrust level (§2). - JouleClaw — the reference cascade and frontier-model substrate behind the synthesis fallback (§6).
- Proof / JouleContract — the JCR-1 receipt format the surfacing log seals to (§8).
- joule-ui — the surface a highlight is rendered on; a surfaced highlight is itself a metered, sealed joule-ui interaction.
- WAI — the container for multimodal sensor payloads referenced by a
Cid.