Tools · L3 — ecash & scaling layers

Evaluating an ecash mint

How to judge whether a Cashu or Fedimint mint is trustworthy — the hard solvency signals, the reputation layer, and the one rule that beats all of them.

What this is

Ecash is the lightest layer in the stack — but it is custodial: a Cashu mint or a Fedimint federation holds the underlying Bitcoin and issues bearer tokens against it, on a fractional-reserve trust model. The token is only as good as the mint behind it. Because anyone can stand up a mint, the practical question for an agent (or its operator) is not “is ecash safe?” but “is this mint one I should hold a balance with?” This page is how to answer that — the objective signals, the reputation layer, and the rule that matters more than either.

(The Cashu and Fedimint cards cover what each protocol is and the single-operator-vs-federated trust split; this page is the vetting companion to both.)

The hard signals (objective)

These are the closest thing to proof, and the strongest filter — though, honestly, few mints publish all of them yet, so “does this mint publish proofs at all?” is itself a quality signal.

• Proof of liabilities + proof of reserves (together). A serious mint publishes its mint proofs (all ecash issued) and burn proofs (all ecash redeemed); the difference is its outstanding liability, which must be backed by the Bitcoin it actually holds. A cheating mint can inflate apparent liabilities but cannot fake on-chain reserves — so the two published together are what make solvency checkable rather than trusted. (Two separate works, one per half. The liabilities scheme is callebtc’s gist, “A Proof of Liabilities Scheme for Ecash Mints” (May 2023). The reserves half is a different paper by different authors — Grunspan & Perez-Marco, “Proof of reserves and non-double spends for Chaumian Mints”, arXiv 2306.12783. Don’t cite one as the formalization of the other; they sit on opposite sides of the solvency equation.)
• Keyset rotation / epochs. A reputable mint runs its signing keys in epochs and rotates them, letting holders migrate tokens to the current keyset and retiring the old one. This bounds the auditable window and blocks a mint from quietly inflating the supply from a retired key. (Covered by NUT-02; “NUT” = Notation, Usage, and Terminology — the Cashu protocol specs, nothing to do with Nostr.)
• Trust model. Single-operator (Cashu) means one party can defect or fail; a federation (Fedimint, minimum 4, commonly 4/7/10, capped at 20 guardians under threshold signatures) reduces — but does not eliminate — that risk. All else equal, distributed custody is the higher-robustness option.
• Lightning liquidity / uptime. A healthy mint can pay out instantly; a mint that keeps failing to settle is the leading indicator of distress, well before any formal insolvency.

The reputation layer (social proof)

Since mints work like local “banks,” reputation does a lot of the work. Three kinds of signal exist today:

• Mint registries / review directories. bitcoinmints.com is the primary directory for both Cashu and Fedimint mints — it tracks each mint, its supported NUTs, and user reviews + vouch counts signed from Nostr identities. cashumints.space is a Cashu-focused sister index with per-mint pages and reliability history.
• Nostr web-of-trust. Because reviews are signed by Nostr keys, vetting can run through your own social graph: modern wallets surface a mint as higher-reputation when accounts you already follow vouch for it. Trusted operators broadcasting a mint list (or a warning) is a real signal — with the honest caveat that Nostr identities are free to create, so reviews are sybil-able and a mint can farm a reputation before exit-scamming.
• Objective status boards. Beyond subjective reviews, automated monitors give an empirical read: calle’s Mint Auditor (audit.8333.space) continuously circulates ecash between mints over Lightning and flags any that fail to pay. (The former cashu.live status board is gone — it has served nothing since 2026-05-18 and its backing repo has been unpushed since 2025-12; github.com/callebtc/cashu-auditor is now a 404, so use the running service rather than the repo path.) Uptime and pay-out behavior are harder to fake than star ratings.
• In-wallet discovery + warnings. You don’t have to chase this down manually: wallets like ZEUS ship a Discover Mints view — which reads NIP-87 mint-recommendation events (kind 38000) off Nostr, filtered to the npubs ZEUS follows, to everyone, or to accounts you name, and points you to bitcoinmints.com to read the reviews yourself. Note what that means for sybil-resistance: the filter is your own follow graph, not a registry’s moderation. Separately, the Bitcoin Design Guide recommends warning a user who pastes a mint with zero vouches or a history of dropping offline.

The rule that beats all of them: don’t concentrate

Even a well-vetted mint can fail, and no proof is perfect. The only fully reliable mitigation is don’t keep meaningful balance in any single mint — spread funds across several reputable ones and hold small operational amounts. ZEUS’s Evan Kaloudis has described an “automated bank run” — guiding users to five or six reputable mints and rebalancing between them — but as a stated intention (Bitcoin Magazine, May 2025, future tense), and as of August 2026 it is not shipped: the phrase appears nowhere in the ZEUS codebase. What ZEUS does ship addresses the same concentration risk from the other end — an automatic sweep to self-custody above a per-mint threshold, which moves value out of ecash entirely rather than spreading it across mints. Plan for the shipped behaviour, not the announced one.

Gotchas

• Proofs are aspirational, not universal. The PoL/PoR tooling exists, but most live mints don’t publish full proofs yet — treat a mint that does as notably more trustworthy, and don’t assume one that doesn’t is insolvent (most aren’t), just unverified.
• Reviews are gameable. Vouch counts and Nostr reviews are a starting filter, never a solvency guarantee; weight your own web-of-trust over raw counts.
• A rating is a snapshot. Uptime and reputation can change fast; for anything beyond a small float, re-check rather than trusting a once-good mint indefinitely.

Links

• bitcoinmints.com — primary Cashu + Fedimint mint registry (reviews / vouches / NUTs). · cashumints.space — Cashu-focused index. · audit.8333.space — calle’s Mint Auditor (live). (cashu.live is dead — no response since 2026-05-18.)
• Proof-of-liabilities scheme: callebtc’s gist “A Proof of Liabilities Scheme for Ecash Mints” (May 2023). For the reserves half — a separate work by different authors — see Grunspan & Perez-Marco, “Proof of reserves and non-double spends for Chaumian Mints”, arXiv 2306.12783 (note the authors’ own published erratum on §3.1). · audit.8333.space — autonomous mint auditor (the old callebtc/cashu-auditor repo is a 404).
• Protocol context: Cashu · Fedimint · the NUTs specs at cashubtc.github.io/nuts.