Notes / Audits and snapshots
Source: chains/zcash/data/scored_targets_2026-09-17.md in the repository, rendered here without changes. Source file
Different in kind from this project's other three 2026-09-17 promotions (Solana, Hyperliquid, Tempo): those each reused a numeric mapping already written and tested a day earlier. Zcash's 0 had qualitative rules (sections 4.1-4.5) but no 0-100 formulas -- section 4.6 and 1 are the methodology-test step and the promotion done together in this one pass, not a promotion of prior-day work.
| Target | Key | adminKey | multisig | timelock | oracleAuthority | crossExposure | composite | l1CappedComposite |
|---|---|---|---|---|---|---|---|---|
| Zcash L1 (consensus-rule-change authority) | 0xB8e62E6AeB4C400b6F3e66cAb69ACe55439035c3 (synthetic, keccak256("zcash:mainnet:l1") per METHODOLOGY.md section 4 target types -- no address controls it) | 20 | 50 | 35 | 100 | 100 | 34 | n/a (this IS the baseline) |
| ZIP 271 one-time lockbox disbursement | t3ev37Q2uL1sfTsiJQJiWJoFzQpDhmnUwYo | 50 | 39 | 0 | 100 | 100 | 32 | 32 |
| ZCG funding stream (FS_FPF_ZCG_H3) | t3cFfPt1Bcvgez9ZbMBFWeZsskxTkPzGCow | 50 | 39 | 0 | 100 | 100 | 32 | 32 |
zec.omft.near (NEAR PoA bridge, EXT, added 2026-09-17) | zec.omft.near | 5 | 18 | 0 | 100 | 100 | 7 | n/a (off-Zcash authority, see below) |
METHODOLOGY.md section 8 left this open ("Identify the NEAR contract that controls zec.omft.near minting and its signer set"). Resolved by reading github.com/near/intents's contracts/poa/factory/src/contract.rs: the factory contract (omft.near) gates minting (ft_deposit) behind a near_plugins access-control role, live-read and cross-checked against a second NEAR RPC (free.rpc.fastnear.com alongside rpc.mainnet.near.org, raises on disagreement -- same hard cross-RPC requirement as Tempo's call2()):
Role::TokenDepositer (mint-capable): bridge-mng.near, a bare NEAR account confirmed to hold exactly ONE full-access key (k=1, n=1) -- and int-mnt-dao.sputnik-dao.near, a Sputnik DAO whose "Requestor" role (the group that can both propose AND execute a mint call) has only 3 members with threshold=1, quorum=0 -- a single vote from any ONE of those 3 passes a call proposal outright. Structurally a 1-of-3, not a real majority-vote DAO for this action.adminKeyScore=5, multisigScore=18 (same admin_key_score_kn/multisig_score_kn ladder already used for the two Zcash funds above).Role::DAO, is held solely by a DIFFERENT, genuinely strong Sputnik DAO (intents.sputnik-dao.near, 5-member council, threshold=[79,100], i.e. 79%) -- but day-to-day minting does not have to go through it, since TokenDepositer is an independent, weaker OR path.l1CappedComposite deliberately NOT computed for this target: the cap models whether a Zcash rule change could override the target's own authority, which does not apply here -- this bridge's mint authority is entirely off-Zcash.Live, re-derived on every run:
(m, n) threshold and full signer pubkeys, decoded from the most recent real spend's revealed redeem script -- cross-checked against a second, independent lightwalletd operator (zcash.mysideoftheweb.com:9067 alongside zec.rocks:443), which must agree exactly or the scorer degrades rather than trusting one source.crossExposureScore) -- a fresh set comparison, not carried over from METHODOLOGY.md 3.2's prior finding.k50/k25 feeding the L1's multisigScore) over the 2,000 blocks ending 10 below the lower tip of two lightwalletd operators, which must agree (tip-relative since the 2026-09-17 second-run methodology test; before that the window was frozen at 3,483,700-3,485,699). Payout addresses are an upper bound on entities, so this input is an upper bound.eu.zec.rocks:443 (12 live samples).Dated, disclosed judgment calls, NOT re-derived every run (documented in scorers.py's own docstrings, with the reasoning for each):
L1 admin-key ladder itself (1 lineage -> 20, 2 -> 45, 3+ -> 65) -- a first-pass calibration with only one live data point to test against today, written down for transparency and future-proofing rather than presented as mechanically derived (same convention Hyperliquid/Tempo's L1 formulas already used for their own untested higher bands).POOL targets (Sapling/Orchard/Ironwood shielded value pools) are DELIBERATELY, PERMANENTLY not given a 0-100 score -- not a placeholder gap. See 1 : the setup-risk framing METHODOLOGY.md 4.1 describes rests on an unfalsifiable historical-trust question (was at least one Sprout/Sapling MPC ceremony participant honest) that no live read can ever confirm, unlike every other score in this project.The ZIP 271 fund and the L1 concentration input were re-derived with an independent reader and matched every published number above. The test changed rules, not today's scores: unresolved P2SH thresholds now take the ladder floor (5/16) instead of 20/20, script suffixes are validated, the concentration window is tip-relative and two-operator. See 0 and METHODOLOGY.md 7.3.
python3 -c "
import sys; sys.path.insert(0, 'chains/zcash')
from scorers import score_all
for r in score_all():
print(r['target'], r['label'], r['compositeScore'], r.get('l1CappedComposite'))
"
Runs in under a minute: a 12-sample lineage check, a 2,000-block coinbase scan, and a small number of GetTaddressTxids windows per fund (4 and 10,009 transactions scanned respectively for the two funds' full history in the range already known to contain their real spends).
Not an audit, not investment advice. All notes ยท Back to the dashboard.