For every protocol catalogued in Comparisons, we studied how it actually works and asked one question: which concrete element could zkCoins evaluate adopting? This section is the result — a deliberate scan of the field for ideas worth importing, with an honest account of what does not transfer.
These are candidates, not decisions
Everything here is a candidate to evaluate, not an accepted design choice. Each idea is rated for Fit (how naturally it maps onto Client-Side Validation + ZK + Bitcoin) and Effort (Low / Med / High / Research-grade). Adoption decisions happen later, per item.
zkCoins' identity is a rare combination — Bitcoin-anchored · Shielded (anonymity set) · Trustless. An element only qualifies if it strengthens one corner without forcing zkCoins to abandon another (no own chain, no own token, no custodian, no trusted hardware). Many of the richest-looking features in other systems fail this test; those are documented under each project's "Doesn't transfer" so we don't cargo-cult them.
The ideas that recur across clusters or carry the most leverage:
View-key hierarchy (spend-vs-view split). Zcash's FVK/IVK/OVK — independently validated by Monero, Firo, and Zano — is the textbook form of the proposed two-key model (operational key sees, spend key spends) and scoped view grants. Fit High · Effort Med.
Note discovery over an untrusted relay. Zcash trial-decryption of key-private ciphertexts + Aztec deterministic note-tags + Penumbra Fuzzy Message Detection — three proven recipes for "find my own bundles without telling the relay which are mine." Directly powers the Nostr delivery + recovery scan. Fit High · Effort Med–High.
Deterministic seed-based recovery. Cashu NUT-13 (derive secrets from seed, batch-scan with a bounded counter) maps almost 1:1 onto "seed re-derives keys + pull-tags." Fit High · Effort Med.
Bitcoin-anchored spent-coin set (S2). The Tornado Cash public nullifier set and the Cashu spent-secret ledger show the interface of the accumulator zkCoins lacks; the open work is anchoring it to Bitcoin rather than a contract or a mint DB. Fit High · Effort Research.
Broadcaster paid from shielded funds. Railgun's pattern — a third party publishes your transaction and is reimbursed from the value moved — is the cleanest way to pay the Bitcoin commitment fee without the spender revealing a funding UTXO. Fit High · Effort Med.
Replicated availability as a precondition. Taproot Assets universe stores + Fedimint t-of-n replication + Plasma's "data loss = fund loss" lesson turn "honest + replicated network" from advice into an enforced replication factor with acknowledgements. Fit High · Effort Med.
A recurring theme across every cluster: the trust model of a system almost never transfers, only its mechanism. The full-history consignments of RGB/Taproot Assets break the anonymity set; the custody of Cashu, the federation of Fedimint/Liquid/Mercury, and the trusted hardware (TEE/SGX) of Shade all contradict trustlessness; the own-chain security budgets of Zcash/Monero/Namada and the EVM exit contracts of Tornado/Intmax2 have no Bitcoin-only analogue; and every transparent overlay (Omni, Counterparty, Runes) leaks exactly what zkCoins hides. The per-cluster pages spell each of these out.