86e8619b71
JCP stops being only specified and starts being consumed. Three sibling standards now gate their real dispatch through Grants, the deterministic runtime, and sealed JCR-1 receipts — each via a small adapter crate with tests proving the gate: - jouleclaw-jcp: GatedTool wraps any jouleclaw-mcp MeteredTool so every MCP dispatch is decided against a signed delegation chain BEFORE a byte reaches the tool. The tool's stable id is the content-addressed resource (a same-named tool on another server is a different audience), the call's energy estimate is debited pre-dispatch, and every decision seals. Tests: a granted tool dispatches and debits; an ungranted tool named by a poisoned result NEVER runs (panicking stub proves it) and the refusal is sealed evidence; the budget stops a runaway loop at exactly the wall. - sandbox-jcp: JcpGate in front of Host::run_and_seal. The Grant binds to the unit's image BY CONTENT ADDRESS — one flipped image byte is a different audience, denied wrong-resource before the executor is ever invoked. The declared energy ceiling is debited from the chain pre-run (enforce before the side effect; energy never refunds), and a revoked grant goes dark. Two receipts per run: the JCP authorization + the sandbox's metered, attested execution, both JCR-1. - joule-code-jcp: GatedStage wraps any cascade Stage so every tier consultation is decided pre-dispatch — the tier is the action, the source tree the content-addressed resource, the tier's energy class the debit. Cheapest sufficient AND authorized: a grant that doesn't name l4 means the remote-frontier escape hatch never fires even when it would close the request; an exhausted budget stops it with a sealed budget-exceeded deny; receipts aggregate per-tier from evidence. Cross-standard path deps inside the monorepo (../../../../jcp/jcp-rs). Each adapter crate is tested in its own workspace; READMEs updated (jcp gains a "Consumers in this repository" section). Co-Authored-By: Claude Opus 4.8 <[email protected]>
4.3 KiB
JouleClaw — Energy-Optimized AI Runtime
The substrate for AI that costs Joules, not tokens.
Capability per Joule, not capability per parameter.
What JouleClaw is
JouleClaw is a pure-Rust, omni-modal AI runtime + harness + agentic layer governed by Transaction Science as an open standard (5th in the open-standards quartet, alongside OpenPay, Smart Byte, EOC, and WAI).
Every operation JouleClaw performs walks an explicit energy-tiered cascade, top to bottom, stopping at the first stage that closes the query:
| Tier | Name | Cost class | What runs here |
|---|---|---|---|
| L0 | Cache | picojoules | Content-addressed cache hits |
| L1 | Lawful | nanojoules | Deterministic, pre-compiled primitive (text/code only) |
| L2 | Embed | sub-millijoules | Matryoshka embeddings + nearest-neighbor + hybrid search |
| L3 | Model | joules | Local SSM / ternary / multimodal / diffusion / etc. |
| L4 | Wire | tens of joules | Remote frontier RPC (escape hatch only) |
L1 (Lawful) is only meaningful for text and code modalities. For image, audio, video and 3D generation, the cascade collapses to L0 / L2 / L3 / L4 — the energy savings come from caching plus dispatch to the cheapest model that meets the quality bar.
Why JouleClaw exists
The mainstream "agentic AI" ecosystem treats inference as the entry point. Every state shift, every tool call, every routing decision runs through a 70-billion-parameter stochastic engine, burning kilojoules to do work that deterministic compute or a hot cache could resolve for microjoules.
JouleClaw inverts the default. Inference is the last resort, not the first. Frozen model weights are less trustworthy than current world state; fresh, provenance-stamped retrieval beats synthesis from training cuts. Every operation is metered in microjoules where the hardware permits it, and in millijoules where it doesn't — the spec is honest about the floor on every platform.
The thermodynamic circuit breaker is the safety mechanism. Set an energy budget; when measured consumption exceeds the budget, the breaker trips and the operation halts. Hallucination loops cannot survive an honest joule ledger.
What's here
| Path | Contents |
|---|---|
SPEC.md |
JouleClaw v1 spec — cascade tiers, energy contract, receipt |
CHARTER.md |
Stewardship pattern (shared across open-standards) |
jouleclaw-rs/ |
Apache-2.0 Rust reference implementation |
spec/ |
Tier specifications, energy provenance, receipt format |
conformance/v1/ |
Signed wire vectors for cross-implementation verification |
Reference implementation
jouleclaw-rs/ is the Apache-2.0 reference implementation in Rust. It
runs on Apple Silicon (M3/M4/M5 with UMA + Metal 4 Tensor APIs), AMD Strix
Halo / Ryzen AI Max, Intel Lunar Lake, NVIDIA Jetson, and discrete
NVIDIA / AMD GPUs.
It is omni-modal: text, code, vision, audio, video, image diffusion, 3D Gaussian splatting, and cross-modal fusion all dispatch through the same energy-tiered cascade.
The MCP tool surface is a JCP consumer: jouleclaw-jcp wraps any
jouleclaw-mcp metered tool so every dispatch is decided by the JCP runtime
against a signed, metered Grant before a byte reaches the tool — the tool's
identity is the content-addressed resource, the call's energy estimate is
debited from the grant's joule budget, and every decision (allow and deny
alike) seals as a JCR-1 receipt. An injected instruction can name an ungranted
tool; it cannot make the gate dispatch it.
Status
v0.1.0 reference implementation in active development. The spec is draft status. Cross-implementation conformance vectors will land with v1.0.0.
License
- Code: Apache-2.0 (see
LICENSE) - Spec text: CC-BY-4.0 (see
spec/LICENSE)