Skip to main content

​
ZK Proof-Artifact Roadmap

This roadmap is research and testnet only. Nothing here describes a live or imminent production capability. There is no production custody, no mainnet deposits, no yield, and no quantum-proof guarantee. The SP1 smoke lane is not production proving, and no on-chain verification exists unless and until it is actually implemented and independently reviewed.
A staged plan for evolving the post-quantum (PQ) verifier and ZK/SP1 proof-artifact work, based on the public reference repo Wallet-Wall/walletwall-vault (current public main: walletwall-vault@0.8.5, SHA 6462c10). It exists to keep the next steps sober and sequenced — not to assert readiness. The WalletWall app consumes only local, read-only evidence shapes from this work. It does not run proving, does not verify on-chain, and does not import prover/verifier code at runtime. See PQ Evidence, Testnet Rehearsal & ZK/SP1 Status and Vault Boundaries & Disclosures.

​
Current state (as of walletwall-vault@0.8.5, baseline walletwall-vault@0.8.1)

ComponentStateNotes
TS / open PQ verifierPure ML-DSA-65 (FIPS 204) verification boundaryOff-chain, read-only; boundary guards merged (public #60)
Evidence artifact / schemawalletwall.pq-verifier.v1, hash-only result shapeDeterministic; carries only keccak256 hashes (public #61)
SP1 smoke laneResearch smoke signal onlyCI lane proving the toolchain builds — not production proving (public #62)
ZK/PQ status matrixDisclosure matrix of what is / isn’t provenSeparates PQ evidence, rehearsal, and ZK status (public #63)
ACVP vector coverageBroader known-answer coverage addedML-DSA-65 verifier coverage expanded (public #64, @0.8.2)
Reproducible proof artifactExample publishedCanonical inputs, expected hashes, and reproduce command — research artifact only, not a production proof (public #65, @0.8.3)
Hosted verifier demoFeasibility spike onlyDev/testnet spike: accepts a test vector and returns a hash-only result; no app runtime call (public #66, @0.8.4)
On-chain verifierFeasibility study onlyGas/calldata/precompile/threat-model document — not an implementation; on-chain ML-DSA verification does not exist (public #67, @0.8.5)
Gated SP1 proving / e2eNot implementedNo end-to-end proof generation is wired or claimed

​
Stages

Each stage is independent and gated. None implies the next is funded, scheduled, or safe to ship.

​
Stage 0 — Read-only evidence (live)

The app shows hash-only PQ verifier evidence and testnet rehearsal metadata as separate gates. This is the only live app capability. No proving, no on-chain verification, no custody.

​
Stage 1 — Broaden ACVP vector coverage ✓ done in public repo

Broader known-answer / ACVP test-vector coverage for the ML-DSA-65 verifier was added in public PR #64 (@0.8.2). Increases confidence in the off-chain verifier; changes nothing about on-chain or production status. No app change required.

​
Stage 2 — Reproducible proof-artifact example ✓ done in public repo

A reproducible, independently re-verifiable proof-artifact example (inputs, expected hashes, and a reproduce command) was published in public PR #65 (@0.8.3). This is a research artifact, not a production proof and not a custody signal. The app may display a reference link and expected hash (read-only) — no prover, verifier import, or raw proof blob.

​
Stage 3 — Gated SP1 proving / e2e (research, not yet built)

Wire end-to-end SP1 proving behind an explicit research gate, clearly labelled as not production proving. Output remains a research artifact; the app continues to display read-only evidence only. This stage is not yet built.

​
Stage 4 — Hosted verifier demo spike ✓ done in public repo (feasibility only)

A hosted verifier spike was completed in public PR #66 (@0.8.4): feasibility/dev service that accepts a test vector and returns a hash-only result. The app does not call this service at runtime. This is docs/spike only — no custody, no mainnet, no user funds.

​
Stage 5 — On-chain verifier feasibility ✓ done in public repo (study, not build)

An on-chain verifier feasibility study (gas, calldata, precompile assumptions, threat model) was authored in public PR #67 (@0.8.5). This is a document, not an implementation. On-chain ML-DSA verification does not exist until separately built and audited.
For Wallet-Wall/walletwall-vault, sequenced and sober. Stages 1, 2, 4, and 5 are done in the public repo; stage 3 remains unbuilt:
  1. More ACVP vectors — broaden known-answer coverage for ML-DSA-65. ✓ done (public #64, @0.8.2)
  2. Reproducible proof-artifact example — publish inputs + expected hashes + a reproduce command. ✓ done (public #65, @0.8.3)
  3. Hosted verifier demo spike — feasibility spike only; testnet/dev, no custody. ✓ done (public #66, @0.8.4)
  4. On-chain verifier feasibility doc — a study (gas/calldata/threat model), not an implementation. ✓ done (public #67, @0.8.5)
  5. Stage 3 — Gated SP1 proving / e2e — wire end-to-end proving behind an explicit research gate, clearly labelled as not production proving. Not yet built. Requires its own design review.
  6. Audit / independent review — before any stage moves toward production.
None of the above introduces custody, signing, mainnet deposits, yield, or a production protection claim.

​
Explicit non-goals

This roadmap does not plan, promise, or imply:
  • production custody of funds,
  • mainnet deposits or withdrawals,
  • yield, interest, or returns,
  • a quantum-proof or quantum-secure guarantee,
  • on-chain ML-DSA verification (none exists today),
  • production SP1 proving (the smoke lane is research only),
  • any runtime dependency of the app on the public vault repo.

PQ Evidence, Rehearsal & ZK/SP1 Status

How the app surfaces these as separate, read-only gates today.

Vault Boundaries & Disclosures

Full custody, testnet, and quantum-resistance boundary statement.

walletwall-vault (public repo)

The public reference repo for contracts, the verifier, and proof artifacts (current: @0.8.5).