What it is. Abelian is a private digital-cash network where you can choose to keep a payment fully hidden, fully open like Bitcoin, or open only to an auditor you hand a viewing key.
What we found. Its live network was already built to survive a future quantum computer, so it skips the upgrade scramble most chains still face, but the whole defence rests on one in-house method that no outside expert has reviewed.
Why it matters. If that single method ever proved flawed there is no second line of defence behind it, and with no independent review there is no early warning to catch the problem first.
Abelian's L1 mainnet has signed close to all transactions under lattice primitives since 17 April 2022, with no classical scheme present: a custom signature inspired by CRYSTALS-Dilithium (ML-DSA family, FIPS 204), a CRYSTALS-Kyber-inspired KEM (ML-KEM family, FIPS 203), and the pqringct / pqringctx linkable ring signatures over LWE / Ring-LWE. The single fact a cryptographer must weigh: this is pure-PQ on one cryptographic family (lattice) with no hash-based or code-based signature fallback, the signature and KEM are custom 'inspired by' constructions rather than verbatim FIPS 204/203 with no third-party audit of pqringct/pqringctx located, so Gate 1a-Sig (no classical+PQ hybrid) and the lattice-monoculture cap both fire (QRI <= 60), and the wABEL bridge plus custody and infra tiers name no PQC vendor commitments, holding migration stage at 3.
Summary
Abelian's L1 mainnet, live since 17 April 2022, signs close to all transactions under lattice primitives with no classical fallback: a custom signature inspired by CRYSTALS-Dilithium (ML-DSA family, FIPS 204), a CRYSTALS-Kyber-inspired KEM (ML-KEM family, FIPS 203), and the pqringct / pqringctx linkable ring signatures over LWE / Ring-LWE. Privacy is the chain's function. It runs a fully-private mode (sender, receiver, and amount hidden by lattice ring signatures plus commitments), a Bitcoin-equivalent pseudonymous mode, and an accountable mode opened by a shareable view key, with one-time addresses that change every use. Three facts hold the score. The deployed PQ posture rests on one family, lattice (LWE/Ring-LWE), with no hash-based or code-based signature fallback, so the lattice-monoculture cap fires. It is pure-PQ with no classical+PQ hybrid signing, which Gate 1a-Sig requires for the top stage, so both ceilings cap QRI at 60. The signature and KEM are custom 'inspired by' constructions, not verbatim FIPS, with parameter sets in an unparsed whitepaper PDF and no third-party cryptographic audit of pqringct/pqringctx located. The wABEL bridge, custody, and infra tiers name no PQC vendor commitments, holding migration stage at 3. The QDay layer-2 (zk-SNARK rollups) is on testnet; its mainnet date is unconfirmed from a primary source.
What the gates say
- Gate 1a, Hybrid signature: FAIL , Abelian is pure post-quantum by design: the L1 consensus/account path signs solely under a custom lattice-based scheme inspired by CRYSTALS-Dilithium / ML-DSA per FIPS 204, plus the pqringct/pqringctx lattice-based linkable ring signature. There is no classical+PQ AND-composition (2-of-2) or OR-composition (1-of-2) with a published EUF-CMA-preservation combiner - there is no classical scheme on L1 to compose with. CORRECTION (verified against AIP-0011 v005 source): AIP-11 derives ONLY the Abelian quantum-resistant Account-Root-Seeds (CoinSpKeyRootSeed, CoinSnKeyRootSeed, CoinDetectorRootKey, CoinVKeyRootSeed) from a BIP-39-style 24-word mnemonic via the KMAC256 PRF KMAC256(key, input, 512, 'ABELIANPRF'); it does NOT derive any ECDSA/secp256k1 legacy account on L1. The dual legacy-plus-quantum-resistant account derived from one mnemonic is a QDay L2 Phase 2 design (whitepaper Section 5, unconfirmed/future), and even there the KMAC256 KDF derives only the quantum-resistant account while the legacy account follows the standard EVM/BIP-39 path. Either way there is no hybrid signature combiner co-signing a single transaction. This is the canonical pure-PQ replacement that Gate 1a-Sig disqualifies, even though Abelian arrived there via a PQ-native architecture rather than a plug-and-play swap. Consequence: QRI cap 60, Stage cap 4.
- Gate 1a, Hybrid KEM: FAIL , Abelian uses a CRYSTALS-Kyber-inspired KEM (ML-KEM per FIPS 203) for key encapsulation/encryption underlying its ring signature and ZKP schemes, but as a pure-PQ KEM with no documented classical+PQ hybrid combiner preserving IND-CCA. The p2p transport-layer cipher suite on port 8666 is undocumented in public material, so no hybrid TLS / shared-secret-concatenation construction (draft-ietf-tls-hybrid-design or equivalent) is evidenced. Either reading fails Gate 1a-KEM: pure-PQ KEM without classical fallback, or undocumented classical transport with no hybrid PQ KEM in scope.
- Gate 1b, Commit-to-hash: COND , no OR-composition / 1-of-2 hybrid declared at L1; Abelian is single-scheme lattice-PQ at the signature layer
- Gate 2, Evidence reconstruction: PASS , -WITH-GAPS (most sub-scores reconstructible in 48h from public artifacts: pqabelian.io, community.pqabelian.io docs/FAQ/release-history, github.com/pqabelian source repos, QDay whitepaper page, BlockchainReporter. GAP: the exact lattice parameter sets and security proofs live only in the whitepaper PDF (binary, not parsed this session) and the QDay mainnet launch block/date is not confirmed from a primary on-chain or foundation-announcement source. No sub-score is fully voided, but Dim 1 1a/1d and Dim 5 5a carry parameter-set / mainnet-date uncertainty reflected in the +-5 CI.
- Gate 3, Primitive naming: PASS , primitives named: custom lattice signature inspired by CRYSTALS-Dilithium / ML-DSA FIPS 204; CRYSTALS-Kyber-inspired KEM / ML-KEM FIPS 203; pqringct / pqringctx lattice-based linkable ring signature over LWE/Ring-LWE; KMAC256 per NIST SP 800-185 based on SHA-3/Keccak; ABEL-ETHash (Ethash variant) and ABEL-Nakamoto (Nakamoto PoW variant); zk-SNARKs adapted from Polygon ZK Rollups on QDay L2. Caveat noted in 1a: the lattice signature is CUSTOM 'inspired by' Dilithium, not a verbatim FIPS 204 deployment.
Burn-vs-rescue policy on file
Declared option f, Undeclared (no quantum-vulnerable-address freeze policy needed; no forward-looking sunset published). Abelian has no classical-key-freeze policy because it has no quantum-vulnerable L1 signature keys to freeze - the L1 has signed under a lattice-based PQ-conjectured scheme since genesis (17 April 2022). The 'burn-vs-rescue' question (forced migration of legacy Shor-vulnerable address spends) is therefore largely moot at L1. The one related migration event is the deprecation of legacy pseudonymous wallets (Desktop Wallet Legacy now marked obsolete) in favor of MLP (Multi-Layer Privacy) wallets with one-click migration, but that is a privacy-format migration, not a quantum-defence freeze. No forward-looking sunset of any L1 primitive has been published. Scored as Undeclared (f) for the policy-coverage sub-field. NOTE: a separate, genuine quantum-defence mechanism DOES exist at QDay L2 - an emergency DAO halt mechanism (whitepaper Section 4.3): on detection of a quantum attack the QDay DAO votes (members prompted to vote within 30 minutes) to halt rollup execution for 12 hours, with attacker accounts frozen via smart-contract call - which is scored under Dim 7 7e, not as a burn-vs-rescue policy.
Seven dimensions
Each dimension scores 0–100 internally; the weighted roll-up produces the QRI.
1 Cryptographic Exposure weight 12% 56 / 100
Abelian publishes a named primitive inventory across the L1 (lattice signature inspired by Dilithium/ML-DSA, Kyber-inspired KEM/ML-KEM, pqringct/pqringctx ring signatures over LWE/RLWE, KMAC256 KDF) and QDay L2 (zk-SNARKs). Deduction: the L1 signature/KEM are CUSTOM 'inspired by' constructions, NOT verbatim FIPS 204/203, and the exact parameter sets (the equivalent of ML-DSA-44/65/87) live only in the whitepaper PDF (binary, unparsed this session). Transport-layer cipher suites on p2p port 8666 are undocumented. Inventory is named but parameter-set-incomplete.
Custom lattice-based signature scheme inspired by CRYSTALS-Dilithium (ML-DSA per FIPS 204, finalized 2024-08-13) - Abelian L1 mainnet signing, live since April 2022; exact parameter sets in unparsed whitepaper PDF · Custom CRYSTALS-Kyber-inspired KEM (ML-KEM per FIPS 203, finalized 2024-08-13) - key encapsulation/encryption underlying ring signature + ZKP schemes · pqringct (original) and pqringctx (enhanced) - lattice-based linkable ring confidential transaction schemes (LWE / Ring-LWE hardness), open-source in Go · LWE / Ring-LWE hardness assumptions (QDay whitepaper Section 2.1) · KMAC256 (NIST SP 800-185, based on SHA-3/Keccak) - AIP-11 key-derivation PRF: KMAC256(key, input, 512, 'ABELIANPRF') · ABEL-ETHash (Ethash / Ethereum 1.0 PoW variant, GPU-friendly) - PoW consensus hash · ABEL-Nakamoto (Bitcoin Nakamoto PoW variant, ASIC/FPGA-friendly) - dual PoW since Aconcagua Fork (block height 464,000, ~Q1 2026) · zk-SNARKs (adapted from Polygon ZK Rollups) - QDay L2 Phase 1, classical (non-PQ), anchored to PQ L1 The entire L1 signing + KEM surface is lattice-PQ (no Shor-breakable signature primitive at L1 consensus or transaction layer), a structurally strong 1b posture. Deductions: lattice-confidence caveat plus the CUSTOM-vs-verbatim-FIPS uncertainty on the signature/KEM; KMAC256 / ABEL-ETHash / ABEL-Nakamoto are Grover-weakened hashes; and QDay L2 Phase 1 uses classical zk-SNARKs (Shor-break) anchored to the PQ L1 - the L2 proof layer is not itself PQ-conjectured-safe.
Custom Dilithium-inspired lattice signature (ML-DSA family)→ PQ-safe lattice (Grover-caveat; lattice-confidence discount per Change 2; CUSTOM-construction caveat - not third-party-audited)pqringct / pqringctx lattice-based linkable ring signature (LWE/Ring-LWE)→ PQ-safe lattice (Grover-caveat; lattice-confidence discount; custom-construction caveat)Custom Kyber-inspired KEM (ML-KEM family)→ PQ-safe lattice KEM (Grover-caveat; lattice-confidence + custom-construction caveat)KMAC256 (SHA-3/Keccak, SP 800-185)→ Grover-weaken (256-bit -> ~128-bit security; used for KDF, not signing)ABEL-ETHash (Ethash variant, PoW)→ Grover-weaken (PoW hashpower halving under Grover; not a signing primitive)ABEL-Nakamoto (Nakamoto PoW variant)→ Grover-weaken (PoW; not a signing primitive; team-claimed quantum resistance not independently reviewed)zk-SNARKs adapted from Polygon ZK Rollups (QDay L2)→ Shor-break (classical zk-SNARK proof system; QDay Phase 1 is classical, anchored to PQ L1 - L2 proofs are not PQ-conjectured-safe)
Exactly ONE PQ cryptographic family is deployed at L1: lattice (LWE / Ring-LWE) - the Dilithium-inspired signature, the Kyber-inspired KEM, and the pqringct/pqringctx ring signature are all lattice. Per the 1c rubric, '1 lattice-only = 5 (hard cap).' No hash-based or code-based SIGNATURE fallback is deployed (KMAC256 is a KDF hash, not a signature family; ABEL-ETHash/ABEL-Nakamoto are PoW hashes). This is the precise lattice-monoculture single-point-of-failure the Family-Diversity Cap targets, and it fires (QRI <= 60).
The L1 signature is inspired by CRYSTALS-Dilithium / ML-DSA (NIST Categories 2/3/5 by parameter set) and the KEM by CRYSTALS-Kyber / ML-KEM (NIST Categories 1/3/5), but because the schemes are CUSTOM 'inspired by' constructions rather than verbatim FIPS 204/203 with published parameter sets, no NIST security category can be definitively mapped from public sources. The whitepaper PDF (unparsed) would be required to confirm whether the deployed parameters map to a specific NIST category. Partial credit: the scheme is in the lattice family NIST standardized, but the exact category is unconfirmed.
Library provenance: first-party open-source implementations in Go - github.com/pqabelian (abec node, pqringct, pqringctx, abewallet) plus a Java SDK (abelian-sdk-java-v2). The lattice schemes have a multi-year production track record since the April 2022 mainnet. Deductions: NO third-party cryptographic audit of pqringct/pqringctx or the QDay node architecture was located (the QDay node-level pq_-verification design is explicitly unaudited per public sources); no machine-checked formal verification (Formosa/EasyCrypt-class) is cited; constant-time / KyberSlash-class timing-leak posture is not independently validated in public; the scheme is cryptanalytic Tier 3-4 (lattice family is NIST-standardized = Tier 3, but a CUSTOM derivative not subject to the NIST process leans toward Tier 4 'research-grade' confidence). The 10% lower-tier discount consideration applies to the custom-derivative status.
2 Quantum Recovery Exposure weight 10% 72 / 100
Forge. Abelian addresses are one-time: 'your Abelian wallet address changes every time you use it,' and spends use a lattice-based linkable ring signature (pqringct/pqringctx). Revealing an address/public key on-chain does NOT enable Shor-based key recovery, because the underlying scheme is lattice (LWE/RLWE), not ECDSA/Ed25519. This combines a PQ-safe signing primitive with one-time-address hygiene, near-eliminating active-key forge exposure. Deduction: the scheme is a CUSTOM lattice construction with no third-party audit located, so the security margin against an adaptive adversary is not externally validated.
Forge. Dormant/cold Abelian balances sit on lattice-based (LWE/RLWE) address schemes that are Shor-resistant by construction - unlike Bitcoin's lost Satoshi-era P2PK or dormant Ed25519 balances, Abelian cold funds are not forgeable by a future CRQC at the signature layer. No public on-chain analytics for Abelian's UTXO set exist (no equivalent of Bitcoin dormant-balance studies), so the quantum-exposure floor cannot be independently quantified, but the cryptographic exposure is structurally low. Deduction: custom-construction / no-audit caveat and the absence of independent on-chain dormant-balance data.
Forge. Historical Abelian L1 transaction signatures are lattice-based (LWE/RLWE) and remain unforgeable post-Shor; signed spends and ring signatures rest on lattice hardness, not on a Shor-breakable curve. Deduction: the lattice-confidence caveat for very-long-horizon validity, the custom-construction / unaudited status, and KMAC256 internal Grover-weakening trimming margins slightly. (QDay L2 zk-SNARK proofs are classical/Shor-break, but those are L2 proof validity, scored under Dim 4/5, not L1 signature longevity.)
Decrypt (HNDL). The p2p transport-layer cipher suite on port 8666 is undocumented in public material; no hybrid PQ KEM at the transport layer is evidenced, so harvested transport ciphertexts could be decryptable once a CRQC breaks any classical key exchange in use. Scored 9/10 - near the top of the privacy-focused-profile 10-point cap (vs 25 for non-privacy) - because Abelian DOES use a Kyber-inspired (ML-KEM) KEM within its protocol stack (ring signature / ZKP layer), a concrete PQ KEM with a named lineage, even though it is not documented as a hybrid combiner and the transport layer specifically is unconfirmed.
Decrypt (HNDL) - privacy-focused-chain sub-score. Abelian's fully-private note payloads (hidden sender, receiver, amount) are protected by lattice-based ring signatures + commitments and a Kyber-inspired (ML-KEM, FIPS 203 lineage) KEM, NOT by Shor-breakable ElGamal/ECIES/KZG encryption - so a harvested fully-private note is NOT trivially decryptable once a CRQC arrives. This is structurally stronger than EC-based shielded chains (Zcash Orchard, etc.). Scored 10/30 (not higher) because: (a) the KEM is a CUSTOM Kyber-inspired construction with no third-party audit located, so 'pure-PQ KEM with documented migration path' (15) is not fully evidenced; (b) the rubric reserves 15 for documented pure-PQ KEM and 20 for PQ-KEM-plus-historical-note-re-encryption, and no historical-note re-encryption/rotation is documented; (c) legacy pseudonymous (pre-MLP) note traffic had Bitcoin-level visibility and its HNDL exposure is unquantified; (d) memo-encryption status is unconfirmed. Credited at the 'hybrid/PQ-KEM-deployed-for-note-ciphertexts but unaudited and no historical rotation' level.
3 Metadata, Anonymity & Confidentiality weight 25% 62 / 100
Anonymity. Abelian's fully-private tier hides sender address, receiver address, AND amount using lattice-based linkable ring signatures + commitment schemes (pqringct/pqringctx) - a Monero/Zcash-class shielded transaction graph, not a transparent pseudonymous ledger. Three tiers exist: fully-private (graph hidden), pseudonymous (Bitcoin-equivalent, graph visible), and accountable (private to public, view-key shareable). Deduction: tier choice is user-controlled, so pseudonymous-tier transactions remain graph-visible; the % of mainnet traffic that is fully-private vs pseudonymous is not publicly reported, so the realized anonymity set is unquantified.
Anonymity. Abelian is an independent PoW network (ABEL-ETHash + ABEL-Nakamoto) with GPU/ASIC mining reported from 32 countries (team FAQ) and a dedicated ABEL-Nakamoto pool (abel-nakamoto.xyz); it does not depend on an Infura/Alchemy-class RPC oligopoly, so top-3 RPC concentration is structurally low. Deductions: exact node/RPC shares are unpublished, the validator/node metadata-retention policy (IP, timing, client fingerprint) is undeclared (zero credit on that component per 3b rubric), and the 32-countries claim is team-authored without independent on-chain confirmation.
Anonymity. The QDay Bridge targets cross-chain transfer to Tron, Ethereum, BSC, Polygon and other EVM chains; bridge endpoints on those classical chains use ECDSA-based wallets, and the wABEL bridge is operated by a TSP-certified institution union (centralized custodial bookkeeping). A passive observer can correlate source-to-destination at the custodial bridge and at exchange on/off-ramps (MEXC, BitMart, XT.COM). Bridge-side privacy is not addressed in public docs. Moderate-exposure score: the L1 itself is shielded, but the bridge and CEX surface are fully correlating.
Confidentiality. This is an Abelian strength relative to EC-based privacy chains. Abelian's privacy primitives are lattice-based (pqringct/pqringctx ring signatures + commitments, Kyber-inspired KEM) - there are NO discrete-log/EC privacy primitives (no ElGamal, no EC ring signatures, no EC zk-SNARKs at the privacy layer), so Shor does NOT retroactively de-anonymize Abelian's fully-private history at the cryptographic layer the way it would for an EC-based shielded chain. Deduction: the custom-construction / no-third-party-audit caveat means the de-anonymization resistance is not externally validated; and legacy pseudonymous (pre-MLP) transactions were never shielded in the first place.
Anonymity. The lattice-based linkable ring signature scheme (pqringct/pqringctx) provides a cryptographic anonymity set, hiding the true spend within a ring. The current Desktop Wallet Pro (MLP) documentation states support for up to 5 fully-private recipients and 100 pseudo-private recipients per transaction. This is a cryptographic shuffle/mix-by-ring-signature mechanism with a named spec, scored above commit-reveal (10) as a cryptographic shuffle, but below the 15 'cMix-class >=3 independent mix nodes' / 20 'structural mix-network with cover traffic' tiers because the ring-size anonymity set is small and there is no documented mix-node network, precomputation, or cover traffic. Named mechanism: lattice-based linkable ring signature with bounded ring size.
Confidentiality - privacy-focused-chain sub-score; X_priv computed here. Encrypted note payloads on Abelian are protected by lattice-based commitments + a Kyber-inspired (ML-KEM, FIPS 203 lineage) KEM rather than Shor-vulnerable ElGamal-on-EC / ECIES / KZG encryption, so the content-payload shelf life is NOT bounded by Shor breaking a classical curve. Per the 3f rubric this sits between 'hybrid PQ note encryption deployed on mainnet for new notes' (15) and lower tiers: scored 12 because the PQ note encryption IS deployed on mainnet (live since April 2022) for fully-private notes - but (a) it is a CUSTOM Kyber-inspired KEM with no third-party audit located, (b) historical-note re-encryption/rotation (the 20-point tier) is NOT documented, and (c) legacy pseudonymous (pre-MLP) note content was never encrypted and its retention/HNDL exposure is unquantified. X_priv is long-shelf-life-PROTECTED for fully-private MLP notes, unprotected for legacy pseudonymous traffic.
4 Migration Architecture weight 12% 55 / 100
Abelian's roadmap names 'Blanc Fork supports post-quantum cryptographic agility' as a PLANNED milestone, using the exact term 'cryptographic agility' - but Blanc Fork has no confirmed date or technical specification in public sources and is not yet live. As of Q2 2026 the most recent confirmed consensus fork is the Aconcagua Fork (block height 464,000, ~Q1 2026, dual PoW), and that was a PoW-algorithm change, not a signature-scheme version switch; an earlier MLP/AUT fork (block height 300,000) is also confirmed. The L1 has no documented in-protocol versioned-signature-type mechanism (EIP-7702/8141-style) to switch algorithms without a hard fork. Partial credit for stated intent + a delivered fork track record; deduction because the agility mechanism itself is announced-not-implemented and the rubric requires a verifiable production instance.
AIP-0011 v005 (Status Active, Created 2024-11-21) defines deterministic per-user derivation of the Abelian quantum-resistant Account-Root-Seeds from a BIP-39-style 24-word mnemonic via the KMAC256 PRF KMAC256(key, input, 512, 'ABELIANPRF') - a documented per-user account-derivation path. NOTE (correction): AIP-11 does NOT derive a legacy ECDSA-compatible account; the shared-mnemonic legacy-plus-quantum-resistant dual-account design is a QDay L2 Phase 2 feature (whitepaper Section 5), which is unconfirmed/future. The MLP wallet provides a one-click legacy->MLP migration deployed in the production Desktop Wallet Pro (MLP) and CLI abewalletmlp, with active wallet adoption and exchange MLP-withdrawal support (MEXC, BitMart, XT.COM). One-time addresses (rotate every use) provide native key-rotation hygiene. Scored 9/20 (down from a prior 11): a documented client-layer migration path exists AND is deployed with active wallet traffic, but it is a privacy/account-format migration rather than a PQ-key-rotation primitive specifically; the previously-credited 'ECDSA dual-account from AIP-11' is not supported by the source; and there is no ERC-4337/EIP-7702-class smart-contract-wallet AA framework on the UTXO L1 (QDay L2 is EVM-compatible but its mainnet status is unconfirmed).
Abelian has executed at least TWO confirmed consensus-level hard forks: the MLP / Abelian-User-Token fork at block height 300,000, and the Aconcagua Fork at block height 464,000 (~Q1 2026), which introduced dual PoW (ABEL-ETHash + ABEL-Nakamoto). Coordinated client upgrades are evidenced by the current shipped artifacts: Desktop Wallet Pro (MLP) v2.2.0 (2026-05-07), Desktop Wallet Legacy v5.0.0 (2026-01-12), abec node v3.0.1 (2026-01-01), mobile Abelian Pro v8.1.0 (date not published on the downloads page), with no contested forks documented. Deduction: the most recent consensus fork is recent (a relatively young adversarial-coordination track record), and the abec node-software release history is not listed in an accessible public changelog format.
Abelian is explicitly PURE post-quantum at L1 and cannot ship a classical+PQ hybrid signature composition because it has no classical L1 scheme to compose with. The QDay L2 Phase 2 design (whitepaper Section 5, unconfirmed/future) places a legacy (EVM/ECDSA) account and a quantum-resistant account under one shared mnemonic, but that is account-coexistence, NOT a hybrid combiner co-signing a single transaction; and AIP-11 on L1 derives only the quantum-resistant account, not an ECDSA one. Under the v3.1 'hybrid mandatory' framing this scores low - not because Abelian is unprepared, but because the dimension rewards classical+PQ co-signing, which Abelian by design does not do. Tension noted: Abelian is penalized here precisely for being PQ-native (further along than the hybrid-transition model assumes), the same structural penalty borne by other PQ-native chains.
Abelian uses STATELESS lattice-based schemes (Dilithium-inspired signature, pqringct/pqringctx ring signatures) at the transaction layer - NOT a stateful hash-based scheme (XMSS/LMS) - so there is no one-time-signature index-reuse / state-exhaustion risk to manage. Per the 4e rubric, 'chains using only stateless schemes score full 15 by default.' (The one-time-ADDRESS rotation is an address-hygiene feature, distinct from stateful-hash OTS-index state.) No Stage-5 90-day stateful-hash operational record is required for Abelian.
N/A. Abelian L1 uses Nakamoto-style PoW (ABEL-ETHash + ABEL-Nakamoto) - block production is hashing, not signature aggregation - so there is no Shor-vulnerable BLS/threshold signature-aggregation path at L1 consensus to declare. The QDay L2 'QDay DAO' is a PoS-over-PoW staking model, but the validator/consensus signature scheme is not specified in the public whitepaper (Dim 4 gap), and QDay mainnet is unconfirmed; it is not documented as using Shor-vulnerable BLS aggregation. Per the 4f rubric, non-aggregating PoW consensus scores N/A and the 20-point weight redistributes across 4a-4e. Dimension score = (6+9+11+3+15)/80 * 100 = 55.0 -> 55.
5 Deployment Execution weight 18% 67 / 100
Abelian L1 mainnet (live since April 2022) signs ~100% of transactions under a lattice-based PQ-conjectured scheme (custom Dilithium-inspired signature + pqringct/pqringctx ring signatures + Kyber-inspired KEM) - there is no classical (ECDSA/Ed25519) fallback on L1. The 5a rubric measures '% of mainnet signing traffic on PQC primitives (hybrid or PURE)' - Abelian is ~100% pure-PQ. Among the highest 5a values in the evaluation set. The Mainnet-Traffic cap therefore does NOT fire. Deductions: the scheme is a CUSTOM lattice construction (not verbatim FIPS) with no third-party audit, and the % fully-private vs pseudonymous traffic split is unreported; the credit is for L1 lattice-PQ signing specifically, NOT for the unconfirmed QDay L2 mainnet.
Lattice-based signature, KEM, and ring-signature verification code is fully merged and in production in the Abelian L1 core node (github.com/pqabelian/abec) and the pqringct/pqringctx libraries (Go) - this is shipped consensus-client PQC code running on a live mainnet, not testnet-only. Near-full credit because, like only a few chains in the set, there is PQ-conjectured signing code in a LIVE mainnet consensus client. Deduction: no third-party audit of the merged crypto code is located, so the implementation correctness is not externally validated.
Abelian L1 is PoW (ABEL-ETHash + ABEL-Nakamoto) - block production is mining (hashing), not validator signing - so there are no 'validator PQC signing keys' at L1 consensus to register; however ~100% of USER transaction signing on the live chain is lattice-PQ. QDay L2 introduces a 'QDay DAO' PoS validator/staking model (min 100,000 QDAY stake), but its validator signature scheme is unspecified and its mainnet is unconfirmed. Scored mid-range: full user-signing-key PQ adoption on the live L1, but no L1 validator signing layer to adopt PQC at and the QDay validator-key story is testnet/unconfirmed.
NOT voided (5a > 0; the v3.1 5a-zero void condition does not apply). Abelian publishes named, dated, verifiable milestones backed by shipped artifacts: mainnet launch (17 April 2022), MLP/AUT fork (block height 300,000), Aconcagua Fork (block height 464,000, ~Q1 2026, confirmed live), HAKO METAL Genesis ASIC for ABEL-Nakamoto mining (product confirmed; exact ship date not pinned to a primary source this session), and continuous wallet releases through 2026-05 (Desktop Wallet Pro MLP v2.2.0, 2026-05-07). Deductions: the milestones are NOT protocol-enforced sunset/flag-day commitments (no governance-bound or consensus-enforced deadline of the EIP-7851 / Optimism-flag-day / BIP-361 type); the Blanc Fork 'cryptographic agility' milestone is undated/unspecified; and the QDay mainnet milestone (whitepaper-targeted Q1 2025) is unconfirmed from a primary source.
Announced ~= shipped at the core: Abelian has signed real mainnet transactions under a lattice-based PQ-conjectured scheme since April 2022, so its central PQC claims are backed by deployed, on-chain primitives - the inverse of narrative-only chains. Announced-vs-shipped ratio ~1.25 (no >1.5 deduction). Deductions reflected: the QDay L2 mainnet is announced/targeted (Q1 2025) but not confirmed launched (testnet since Q3 2024), and 'world's first post-quantum blockchain' / 'world's first quantum-resistant EVM-compatible L2' are unverified primacy claims; external escalation of 'inspired by' to 'based on CRYSTALS-Dilithium' is washing-adjacent. Core deployment honesty is high; forward/L2 claims and primacy framing trim the score.
Abelian's deployment cost is high and only partially disclosed. The signature scheme is lattice-based inspired by CRYSTALS-Dilithium (ML-DSA reference: ML-DSA-44 sig ~2,420 B ~= 38x the 64-byte ECDSA/Ed25519 baseline; larger parameter sets are higher), and the pqringct/pqringctx linkable ring signatures with up to 5 inputs add further per-transaction overhead well above a single signature. The exact per-block/per-tx byte multiplier under Abelian's CUSTOM parameters is NOT disclosed in parseable public material (it sits in the unparsed whitepaper PDF), and Abelian deploys NO SNARK aggregation, data-availability offloading, or batching at L1 to bring the effective multiplier down. No user-laggard decay-rate reporting. Scored 4/20: a non-zero floor because the lattice-ring footprint is the honest cost of strong PQ privacy and the scheme is named, but the raw multiplier is undisclosed-and-large (ring signatures push it above the >38x band) with no aggregation.
6 Supply Chain Vendor Readiness weight 18% 24 / 100
Abelian's first-party wallets are PQ-native by construction: Desktop Wallet Pro (MLP), CLI wallets (abewalletmlp / abewalletlegacy), and the Mobile app all sign with the lattice-based scheme today, all open-source (Go for abec/abewallet; Java SDK abelian-sdk-java-v2). Exchanges MEXC, BitMart, XT.COM confirmed supporting MLP addresses. Scored well above the set norm because the PRIMARY wallet IS quantum-safe rather than carrying a 'PQC roadmap.' Deductions: the wallet ecosystem is first-party-only (no MetaMask/Phantom-class third-party support), and NO hardware wallet (Ledger/Trezor) integration for ABEL is confirmed despite Trezor's PQBD 2025 sponsorship - so the hardware-custody leg is missing.
The wABEL bridge (ABEL L1 -> QDay L2) is to be operated by 'a union of institutions having Trust Service Provider (TSP) certification' - a centralized custodial model with NO named TSP institutions and NO PQC HSM requirement specified. The QDay cross-chain bridge to Tron/Ethereum/BSC/Polygon uses classical ECDSA keys on the non-PQC chain endpoints. No third-party bridge vendor (LayerZero/Wormhole/Axelar/CCIP) is integrated or publishes an Abelian-specific PQC roadmap. Worst tile alongside custodian/infra - drives the Supply-Chain cap (3 of 4 tiles lack named top-3 vendor PQC roadmaps -> Migration Stage <= 3).
No independent institutional custodian (Fireblocks, BitGo, Anchorage) is named for ABEL, and the wABEL bridge custody falls to an unnamed TSP-certified institution union with no documented PQC key-management protocol. ABEL custody today is effectively exchange-internal (MEXC, BitMart, XT.COM, SuperEx, Coinstore), classical and not PQC-roadmapped for ABEL specifically. No custodian-MPC path is documented for Abelian's lattice signature scheme (ML-DSA-family lattice is MPC-amenable in principle, but no Abelian custodian has implemented it). Low tile, contributes to the Supply-Chain cap.
RPC is served by independent full nodes and mining pools (e.g. abel-nakamoto.xyz); no named third-party RPC provider with an Abelian PQC roadmap. NO HSM vendor (Ledger HSM/Thales/YubiHSM/AWS KMS) is documented as supporting Abelian's lattice keys, no QDay validator HSM requirement is specified, and no TEE attestation chain is in Abelian's path. The infra runs PQ signing, but the third-party infra vendors carry no Abelian-specific PQC roadmap commitments. Cross-tile weak-link: with bridge/custodian/infra all low, the dimension is capped by its weakest tiles - this fires the Supply-Chain cap (Migration Stage <= 3).
7 Governance & Coordination weight 5% 51 / 100
Abelian L1 is PoW (ABEL-ETHash + ABEL-Nakamoto) with GPU/ASIC mining reported from 32 countries (team FAQ) and a stated goal that 'the last thing the Abelian project wants is for the majority of computational power to come from the core team itself.' But NO Nakamoto coefficient or miner/pool-share distribution is publicly reported, and the geographic-distribution claim is team-authored without independent on-chain confirmation. QDay L2's PoS 'QDay DAO' (min 100,000 QDAY stake; 50% of QDAY to validators) is testnet/unconfirmed. Modest, undocumented decentralization.
Coordinated client/consensus upgrades since the April 2022 mainnet, with at least two confirmed consensus hard forks (MLP/AUT at block height 300,000; Aconcagua at block height 464,000, ~Q1 2026) and an active wallet/client release cadence (current shipped artifacts: Desktop Wallet Pro MLP v2.2.0 2026-05-07, Desktop Wallet Legacy v5.0.0 2026-01-12, abec node v3.0.1 2026-01-01, mobile Abelian Pro v8.1.0) with no contested forks documented. Deduction: the cadence is steady but not battle-tested under adversarial time pressure, and the high-stakes coordination event (a second cryptographic family or hybrid migration to clear the gate/cap) has not been attempted.
Clear named coordination lead: Abelian's founders are named cryptography academics - Professor Duncan Wong (Founder), Huaxiong Wang, Khoa Nguyen, Guomin Yang - confirmed by multiple sources (BlockchainReporter 2025; PQBD 2025 event Nov 19 2025 where Duncan Wong is quoted), with the Abelian Foundation as the coordinating entity and an AIP (Abelian Improvement Proposal) process on GitHub. Strong, transparent, technically-credentialed leadership. Deduction: no named PQC working-group lead distinct from the founding team, and AIP throughput/approval metrics are not publicly retrievable.
Abelian's entire thesis is proactive PQ defence (PQ-conjectured from genesis, April 2022), a positive structural signal credited modestly. But it has NOT executed a cryptographic migration under live adversarial pressure - there has been no attack-driven emergency fork. The Aconcagua Fork was a planned PoW-algorithm addition, not an adversarial crypto-migration test. A true adversarial-coordination precedent is absent.
QDay L2 documents an explicit emergency halt mechanism (whitepaper Section 4.3): on detection of a quantum attack, the QDay DAO votes on a proposal to halt rollup execution (DAO members are prompted to vote within 30 minutes; on passing, the halt is for ~12 hours, extendable), during which bridge transfers are blocked because they require quantum-resistant rollup-operator signatures; attacker account freezing via a smart-contract call is also possible and is unaffected by the rollup halt. This is a documented design (scored as a rate-limited / governance-halt response mechanism, above honeypot-5 and rate-limit-10 toward the 'tripwire + response' tiers). Deductions: it applies ONLY to QDay L2 (Abelian L1 has NO equivalent documented tripwire), it depends on DAO members voting within the 30-minute window with no automated backup, and QDay mainnet is unconfirmed - so the mechanism is design-stage, not demonstrably live.
X + Y vs Z, when does the math turn against you?
v3.1 demotes the X+Y vs Z timing test to a secondary signal, the headline output is Migration Stage. The timing test still answers the question: can this chain finish migrating before the threat lands?
Verdict
X + Y does NOT exceed Z for Abelian's signature layer - it is among the rare chains already inside the risk window for value protection, because lattice-based signing structurally mitigates the forge threat that drives X+Y>Z on classical chains. Residual risk concentrates on (a) lattice-confidence and the CUSTOM-vs-verbatim-FIPS / unaudited-construction uncertainty, (b) classical / undocumented p2p+KEM transport and classical bridge endpoints (HNDL / Decrypt), and (c) lattice-monoculture single-point-of-failure (no second cryptographic family deployed).
Z-compliance
Inside the compliance window for the dominant value-protection threat: Abelian's lattice-based signing is QV-resistant, so it is not exposed to the NIST IR 8547 2030/2035 disallowance of quantum-vulnerable public keys the way ECDSA/Ed25519 chains are. Open compliance items: (a) the scheme is CUSTOM 'inspired by' FIPS 204/203 rather than NIST-validated, so it does not carry a FIPS certification a regulator could rely on; (b) p2p/KEM transport hybrid status is undocumented; (c) the wABEL bridge endpoints on ETH/BSC/Polygon use classical ECDSA. Founder team is academic, spanning Hong Kong and Singapore institutions (Duncan Wong, CityU HK lineage; Huaxiong Wang, NTU Singapore; Guomin Yang, SMU Singapore; Khoa Nguyen) per public bios; no single named primary jurisdiction for the protocol/foundation confirmed from public sources.
Source-disagreement disclosure
v3.1 requires every chain card to publish material divergences among authoritative sources, plus the delta-QRI under alternative weighting.
Abelian's own GitHub org and community FAQ state the signature scheme is 'inspired by NIST standardized algorithms such as CRYSTALS-Dilithium and CRYSTALS-Kyber'; some third-party descriptions (e.g. CoinMarketCap's ABEL profile) escalate this to the firmer 'the cryptographic backbone of Abelian is built on CRYSTALS-Dilithium and Kyber' (though the same CMC text elsewhere also repeats the accurate 'inspired by' phrasing). 'Inspired by' / 'built on' is not 'verbatim FIPS 204/203.' The exact parameter sets, the reduction (clean reduction to ML-DSA/ML-KEM vs a novel construction), and the security proofs are in the whitepaper PDF (binary, not parsed this session) with no third-party cryptographic audit of pqringct/pqringctx located. This evaluation treats the scheme as a CUSTOM lattice construction with developing third-party-validation confidence (NOT a NIST-validated FIPS implementation), which lowers Dim 1 1a/1d/1e and widens the CI. This disagreement, together with the unconfirmed QDay mainnet date, is the main driver of the +-5 CI: the artifact-count formula treats Dim 1 and Dim 5 as well-evidenced, but the most decision-relevant facts (exact parameter sets, third-party audit, QDay mainnet block) are unverified, so those two dimensions are placed in the 3-4-artifact CI tier rather than the 5+ tier.
The QDay whitepaper roadmap (Section 7) targets 'Q1 2025 Launch of QDay Mainnet and ABEL Staking'; the mobile wallet added 'Support QDay (Layer 2)' on 2025-10-28, suggesting L2 was app-accessible by late 2025. But the exact QDay mainnet launch date/block is not confirmed from a primary on-chain or foundation-announcement source. This evaluation assigns ALL Dim 5 mainnet-traffic credit to the Abelian L1 (live April 2022, lattice-PQ signing), NOT to the unconfirmed QDay L2 mainnet. QDay zk-SNARK rollups are scored as testnet/architecture capability only.
The original Abelian-User-Token (AUT) protocol was introduced at the block-300,000 hard fork (confirmed by the foundation blog dated 2024-07-29, 'Multi-Level Privacy and User Token Protocol at Block Height 300,000', and by AIP-0011's compatibility section referencing AbewalletMLP-v1.0.1 at height 300,000). The later Aconcagua Fork (block 464,000) added 'Cloaked AUT' (a privacy-enhanced AUT variant) plus dual PoW, per the abec-v3.0.1 release notes (dated 2026-01-01: 'Introduced the Aconcagua Fork at block height 464,000, enabled Cloaked AUT and hybrid PoW with DSA... set commit height at block 480,000'). Some secondary mining-config notes conflate 'AUT' with Aconcagua; the primary sources show original AUT at 300,000 and Cloaked-AUT at 464,000. The two-confirmed-consensus-fork track record (300,000 MLP/AUT; 464,000 Aconcagua/Cloaked-AUT/dual-PoW) stands; the Aconcagua commit/enforcement height is block 480,000.
The Family-Diversity cap is written for lattice-monoculture (the SIKE-2022 single-point-of-failure precedent). Abelian's entire deployed PQ posture - signature, linkable ring signature, and KEM - rests on lattice hardness (LWE / Ring-LWE), with no deployed hash-based or code-based signature family as fallback. KMAC256 (SHA-3/Keccak) is used for key derivation and ABEL-ETHash/ABEL-Nakamoto for PoW, but neither is a fallback SIGNATURE family. The cap therefore fires literally (QRI <= 60). Note: this is a lattice-monoculture, which 1c hard-caps at 5 (more fragile than a hash-based monoculture, which 1c scores at 10).
A prior draft of this scorecard stated that 'AIP-11 derives BOTH a legacy ECDSA-compatible account and a new quantum-resistant account from the same BIP-39 mnemonic via a KMAC256 KDF.' Verification against the primary source (AIP-0011 v005, Status Active, Created 2024-11-21) shows this is NOT correct: AIP-11 derives ONLY the Abelian quantum-resistant Account-Root-Seeds (CoinSpKeyRootSeed, CoinSnKeyRootSeed, CoinDetectorRootKey, CoinVKeyRootSeed) via KMAC256(key, input, 512, 'ABELIANPRF'); it contains no ECDSA/secp256k1 component. The 'shared-mnemonic, separate legacy and quantum-resistant accounts' design is specified for QDay L2 Phase 2 (whitepaper Section 5), which is unconfirmed/future, and even there KMAC256 derives only the quantum-resistant account. The prior draft also dated AIP-11 to '2025-01'; the source PDF/markdown is dated 2024-11-21. These corrections were applied to gate sig_hybrid, 4b, and 4d, and lowered 4b from 11 to 9 (the dual-account migration claim was over-credited), reducing Dim 4 from 58 to 55 (band 6 and Stage 3 unchanged).
A prior draft of this scorecard scored sub-score 2d (Encryption confidentiality / HNDL Decrypt) out of a 25-point - then 20-point - maximum, which is the NON-privacy-chain weighting. For the privacy-focused-chain profile the methodology fixes the Dim 2 sub-score maxima at 2a 20 / 2b 20 / 2c 20 / 2d 10 / 2e 30 = 100, with the Decrypt class carried mostly by the new 2e note-ciphertext sub-score. 2d max is corrected to 10 here (score held at 9/10, near the top of the privacy cap because a named Kyber-inspired ML-KEM is in the protocol stack). This raises the Dim 2 dimension score from a mis-normalized 65 (72/110) to the methodology-correct 72 (72/100), and lifts raw QRI from 54 to 55. Band 6 Piloted and Migration Stage 3 are unchanged; the Gate-1a-Sig and Family-Diversity QRI<=60 ceilings remain non-binding on the number (55 < 60).
A prior draft repeatedly dated the Aconcagua Fork (dual PoW ABEL-ETHash + ABEL-Nakamoto, activated at block height 464,000) to 'Q1 2025.' Primary/secondary sources place it around Q1 2026: the Abelian hybrid-PoW Q&A blog is dated 2026-01-23 and the foundation's completion announcement frames it as 'as we step into 2026.' The earlier MLP / Abelian-User-Token hard fork at block height 300,000 is a separate, earlier confirmed consensus fork. Net effect: Abelian has at least TWO confirmed consensus-level hard forks (block 300,000 MLP/AUT; block 464,000 Aconcagua), not one as the prior draft implied; but the Aconcagua track record is younger (~months, not ~15 months). Date corrected to Q1 2026 in 1a/4a/4c/5d/7b/7d; the two-fork track record is reflected in 4c/7b without raising those scores.
Delta-QRI under alternative weighting
Under an alternative weighting that down-weights Deployment (Dim 5) and up-weights Supply Chain (Dim 6) - the axes where Abelian is respectively strongest and weakest - QRI falls (e.g. Dim 5 18% -> 10%, Dim 6 18% -> 26%: QRI ~= 48, Band 5 Prototyped). A weighting that credited shipped-mainnet-PQC more heavily (Dim 5 -> 26%) lifts QRI to ~57, still Band 6 Piloted (capped at 60 by the hybrid-gate and Family-Diversity caps regardless). A weighting that heavily up-weighted privacy (Dim 3 components) would lift Abelian given its strong fully-private/anonymity posture, but the Confidentiality component is bounded by the custom-KEM audit gap. Abelian is sensitive to the Deployment-vs-Supply-Chain split because it is an outlier on both.
Announcement-to-shipped ratio
Announced: 5. Shipped: 4. Ratio: 1.25.
Tag: none, ratio <= 1.5; Abelian has signed real mainnet transactions under a lattice-based PQ-conjectured scheme since April 2022, so its core PQC claims are backed by deployed on-chain primitives, the inverse of narrative-only chains. The shortfall driving the ratio above 1.0: the QDay L2 mainnet (announced/targeted Q1 2025 in the whitepaper) is not confirmed launched from a primary source (testnet since Q3 2024), and the 'world's first post-quantum blockchain' / 'world's first quantum-resistant EVM-compatible L2' primacy claims are unverified. The 'inspired by' (custom) vs verbatim-FIPS distinction is a washing-adjacent risk: external descriptions (e.g. CoinMarketCap) escalate 'inspired by' to 'based on CRYSTALS-Dilithium'.
Peers in the privacy-focused chain profile
9 chains closest to Abelian by Stage then QRI.