feat(client): resolve-before-pay for commitment-bound quotes (ADR-0004)

[grumbach]

ADR-0004: commitment-bound quote pricing — client side.

Implements "the client pays nothing it cannot resolve". Before paying, the client runs the full binding check on every quote and candidate, so a node that overstates its storage to inflate its price is dropped before any on-chain payment instead of after.

What

For each single-node quote and each merkle candidate, the client now:

• verifies the quote's own ML-DSA-65 signature and that it is for the requested content;
• checks the (committed_key_count, commitment_pin) shape, the count cap, and that price == calculate_price(count) by exact recomputation;
• for a bound quote, fully resolves the shipped commitment: parse it, check it is bound to the quoting peer, verify its signature, that it hashes to the quote's pin, and that its key_count matches the claimed count.

Anything unresolvable, withheld, off-curve, or contradictory is dropped before payment — exactly as the storer would reject it. The verified commitments are then forwarded as sidecars in the PUT bundle (single-node and merkle) so storers cross-check synchronously; sidecar blobs are size-capped before parsing. Pricing and commitment verification come from ant-protocol, so client and node never disagree.

Tests

Adds an e2e test (real QUIC nodes + Anvil) proving bound quotes are shipped, priced, fully resolve, and round-trip; the off-curve / forged-commitment / signature / pin / count rejection paths are covered in unit tests.

Release ordering

Part of the coordinated ADR-0004 cutover — publish order evmlib 0.9.0 → ant-protocol 2.3.0 → ant-node / ant-client. Depends on the shared layer (WithAutonomi/ant-protocol#15; evmlib WithAutonomi/evmlib#11 merged, pending publish) and on the node side (WithAutonomi/ant-node#149). Builds today via committed, temporary [patch.crates-io] git deps plus a temporary git dep on ant-node; both are stripped and repinned to the published crates at cutover.

[jacderida]

Testnet analysis — DEV-01 (5h20m, 58 nodes ×15, 30% NAT) vs release 0.2.10 / node 0.14.2 baseline

Ran this client (paired with ant-node #149) on a staging testnet against an identical-topology baseline over a matched 5h20m window.

Clean result — 0 upload failures (2,977 uploads), 0 download failures (662). Client-side chunk retry amplification = 1.000 exactly: all 98,866 stored chunks succeeded on the first attempt (retries_histogram [98866,0,0,0]) — no quorum-walk / resolve-before-pay retry cost observed. No payment rejections node-side.

Worth noting: uploads ran +26–60% slower than the baseline at every file size (e.g. 1GB 179s→255s, 500MB 105s→168s) and downloads +13–26% slower. Store-latency analysis pins the slowdown on the node side (put_rpc p95/p99 rising as node disk fills), not the client — client store p50 stayed flat (~3s) with retry-amp flat at 1.0.

No client-side blockers found.