CoinJoin is a multi-party Bitcoin transaction in which participants pool inputs and receive equal-value outputs, breaking the common-input-ownership heuristic chain-analysis firms rely on for clustering — the principal on-chain privacy-enhancing pattern available to Bitcoin holders. The concept dates to Gregory Maxwell's 2013 bitcointalk post; serious implementations include JoinMarket (2015), Wasabi (zkSNACKs, 2018; WabiSabi 2022), and Samourai's Whirlpool (2019). The privacy landscape changed materially in 2024: in April the US DOJ indicted Samourai's developers and seized the service; in June zkSNACKs voluntarily shut down Wasabi's coordinator and exited the US market. As of 2026 the surviving sovereign tool is JoinMarket, with Sparrow Wallet and a community-maintained Wasabi (v2.8.0, June 2026) continuing to provide CoinJoin against remaining infrastructure. The compliance footprint is real — post-CoinJoin deposits at KYC exchanges sometimes trigger account freezes — and the legal-uncertainty footprint is non-trivial. The privacy benefit is also real: a well-executed CoinJoin meaningfully raises the chain-analysis cost of post-mix tracing.


Why this note matters

CoinJoin is the load-bearing tool in the Bitcoin privacy stack — the one transaction pattern that materially defeats chain-analysis clustering when correctly used. It is the operational answer to the threat catalogued in Address reuse and chain analysis and the principal hygiene for the post-KYC withdrawals flagged in KYC leakage. The 2024 enforcement actions (Samourai indictment, Wasabi withdrawal) reshaped the tool landscape in ways older guides do not yet reflect; the broader regulatory context is treated in Tornado Cash sanctions and the privacy-tool regulatory landscape. Anonymity-set claims marketed by historical implementations are sometimes overstated, the compliance footprint at KYC exchanges is real and jurisdiction-asymmetric, and the operational practice — fees, UTXO management, post-mix discipline — is non-trivial. The defensible position: CoinJoin is a meaningful privacy improvement for holders whose threat model justifies it and who can absorb the operational and compliance friction. For holders without that threat model, address-reuse hygiene plus the tiered-wallet pattern (see Hot vs cold storage) capture much of the practical benefit with much less friction.


What CoinJoin is

CoinJoin is a single Bitcoin transaction with multiple inputs from different participants and multiple outputs of equal denominations going to those participants. From the on-chain view, the transaction looks like:

Inputs:   [Alice's UTXO]  [Bob's UTXO]  [Carol's UTXO]  ...
Outputs:  [0.1 BTC]  [0.1 BTC]  [0.1 BTC]  [0.1 BTC]  ...  [change to each participant]

The critical property: each output of 0.1 BTC could belong to any participant. An external observer (chain-analysis firm) cannot tell which output corresponds to which input without additional information.

The mechanism in compressed form:

  1. Participants register: each participant signals “I want to mix N coins; here’s my input UTXO and a fresh output address.”
  2. A coordinator constructs the transaction: the coordinator assembles the inputs and outputs into a single multi-input multi-output transaction with the equal-denomination structure.
  3. Each participant signs only their input: critically, no participant signs the whole transaction; each participant signs only their own contribution. The Partially-Signed Bitcoin Transaction (PSBT) format makes this practical.
  4. The transaction is broadcast once all inputs are signed.

The trust model: the coordinator cannot steal funds. Each participant’s input only releases funds if it is signed by the input’s owner, and each participant signs only after verifying that one of the equal-denomination outputs is paying their fresh receive address. The coordinator can refuse to coordinate, can deny service, can identify participants if it controls multiple inputs (the Sybil-attack failure mode discussed below), or can collude with chain-analysis firms by recording the input-output mapping it constructs. The coordinator cannot, however, steal coins.

This is the key cryptographic property that makes CoinJoin sovereign: the participant retains full custody throughout. There is no point at which a third party can abscond with funds.

What CoinJoin defeats

The common-input-ownership heuristic (CIOH): in an ordinary transaction, the inputs are presumed co-owned. In a CoinJoin, the inputs are demonstrably not co-owned. Chain-analysis firms can identify CoinJoin transactions (the equal-output structure is distinctive) and apply different clustering rules — they typically do not propagate CIOH through CoinJoin transactions.

The change-detection heuristic: in an ordinary transaction with two outputs, one is identifiable as change. In a CoinJoin with N equal-denomination outputs, change-detection fails — none of the N outputs is identifiable as change.

Forward-tracing through the transaction: the firm can identify that a CoinJoin happened but cannot link a specific input to a specific output with certainty better than 1/N — where N is the anonymity-set size (the number of equal-denomination outputs in the transaction).

What CoinJoin does not defeat

Identification as a CoinJoin participant: the on-chain footprint is distinctive. Anyone observing the chain knows you participated in a CoinJoin. The chain-analysis firm doesn’t need to identify your specific output to flag your wallet as “this wallet has CoinJoined.”

Sybil attacks on the anonymity set: if the coordinator and/or several “participants” are controlled by the same adversary, the apparent anonymity set is smaller than the structural anonymity set. The Sybil adversary knows their own inputs and outputs; they can subtract them from the apparent set. A CoinJoin with N participants where M are Sybils has effective anonymity-set N-M.

Post-mix discipline failures: if you combine your post-mix output with a non-mixed UTXO in a subsequent transaction, the CIOH propagates and the mix is effectively un-done. Post-mix UTXOs must remain isolated until they are independently spent.

Tooling fingerprints: the wallet software that constructs the CoinJoin leaves a fingerprint (which CoinJoin implementation, which version, default fee patterns). The fingerprint doesn’t identify the specific participant but narrows the wallet ecosystem.

Pre-mix linkability: if your input UTXO was derived from a KYC withdrawal, the chain-analysis firm has your identity attached to the input cluster. The CoinJoin breaks forward propagation but does not retroactively erase pre-mix identity.

Off-chain identity-resolution channels: KYC records, surveillance integrations, dark-market data sales — none of these are affected by CoinJoin participation. CoinJoin defeats one specific layer of chain-analysis; it does not provide blanket privacy.

The honest read: CoinJoin meaningfully raises the chain-analysis cost of post-mix tracing but does not provide cryptographic privacy. The defensible expectation is “my post-mix coins are clustering-ambiguous from this point forward,” not “my coins are untraceable.”


Tool history (and what remains in 2026)

The original proposal (2013)

Gregory Maxwell’s August 22, 2013 bitcointalk post introduced the CoinJoin concept. The core insight: Bitcoin transactions can have multiple inputs from different parties without those parties needing to trust each other; signature requirements ensure each input is signed only by its owner. Maxwell’s post described the trust model and the basic transaction structure; the construction was sketched but not implemented.

Maxwell’s role: see Greg Maxwell for the thinker treatment. The Maxwell post is one of the earliest substantive privacy-tool designs in Bitcoin and remains a reference point.

JoinMarket (2015–present)

Adam Gibson (waxwing) and contributors implemented JoinMarket in 2015 as a peer-to-peer market for CoinJoin participation. The model:

  • Takers are users who want to mix; they pay a fee to makers who offer their UTXOs as mix counterparties.
  • The fee is small (typically 0.01-0.05% of mixed amount); the system runs over IRC for coordination and direct peer signing.
  • No central coordinator: any participant can be a maker or a taker; the matching is market-based.

JoinMarket’s properties:

  • Highly sovereign: no centralized coordinator; the maker-taker market is decentralized.
  • Operationally complex: requires running a node, configuring IRC connections, managing maker/taker liquidity.
  • Smaller user base than the Wasabi/Samourai era; the operational complexity is a real barrier.
  • Continues to operate in 2026; was unaffected by the 2024 enforcement actions because of its sovereign architecture.

JoinMarket is the canonical sovereign-CoinJoin tool. For technical holders willing to absorb the operational complexity, it remains the most-private and most-resilient option.

Wasabi Wallet (2018–present)

zkSNACKs launched Wasabi Wallet in 2018, implementing the Chaumian CoinJoin protocol (a Wasabi-specific evolution of Maxwell’s original design) with a coordinator that handled participant matching and transaction construction. Key features:

  • Equal-denomination output structure (0.1 BTC initially; later 0.05 BTC). Participants joined a pool, paid a small coordinator fee plus mining fees, received 0.1 BTC equal-denomination outputs and change.
  • Tor-only coordinator to limit deanonymization via network-level analysis.
  • Chaumian blind signatures to prevent the coordinator from learning input-output mappings.
  • Wabisabi 2022 upgrade introduced credentials-based participation, variable-denomination outputs, and other improvements that increased anonymity-set flexibility.
  • Easy-to-use desktop wallet UX that made CoinJoin accessible to non-technical holders.

Wasabi’s footprint:

  • Largest CoinJoin user base by transaction volume for most of 2018-2024.
  • Made CoinJoin practical for ordinary holders who could not run JoinMarket.
  • Compliance friction increased over time: many KYC exchanges began flagging Wasabi outputs; Wasabi’s reputation among compliance officers became increasingly negative.

zkSNACKs, the company behind Wasabi, made the decision in June 2024 to shut down the Wasabi coordinator service entirely, exit the US market, and discontinue active development of the consumer Wasabi product. Public statements cited the 2024 enforcement-action environment (specifically the Samourai indictment) as the proximate cause. As of 2026, the Wasabi wallet software still exists and can be used with other coordinators, but the canonical zkSNACKs coordinator is offline.

Development did not end with zkSNACKs, however. Wasabi has continued as a community-maintained open-source project, and the v2.8.0 release (June 27, 2026) was its most substantial post-shutdown iteration: peer-to-peer synchronization of compact block filters (removing reliance on a central indexing server — a meaningful sovereignty improvement for sync privacy), an experimental Pay in CoinJoin feature (making a payment as part of a CoinJoin round, concealing input age and change patterns), payment batching, sub-1 sat/vByte fee support, arm64 Linux / Tails / Whonix support, an experimental Scheme scripting language for wallet automation (sandboxed away from keys and funds), signet support, and a bundled-Tor upgrade for forward compatibility ahead of the Tor network’s September 1, 2026 deprecation of older versions.

Samourai Wallet (2015–2024)

Samourai Wallet launched in 2015 as a mobile-first privacy-focused Bitcoin wallet. Its CoinJoin implementation, Whirlpool, launched in 2019 with a distinctive architecture:

  • Pre-mix pools: rather than ad-hoc CoinJoin transactions, Whirlpool used denominated pre-mix pools. A participant deposited into the pre-mix pool; the pool periodically constructed CoinJoin transactions among pool participants; the participant could then withdraw post-mix outputs.
  • Tx0 transactions: the entry-into-pool transaction was a distinctive “Tx0” pattern.
  • Smaller equal-denomination pools (0.001 BTC, 0.01 BTC, 0.05 BTC, 0.5 BTC) to accommodate different user-balance ranges.
  • Stronger structural anonymity-set guarantees than ad-hoc CoinJoin: the pool maintained a sustained set of participants over time, so individual mix transactions drew from a documented anonymity-set.

Samourai’s footprint:

  • Strong user base among privacy-focused mobile-first holders.
  • Distinctive Samourai team culture — visibly adversarial relationship with the regulated-finance world; deliberately confrontational public posture.
  • Increasing legal-uncertainty exposure over 2022-2024, culminating in the April 24, 2024 indictment of Samourai co-founders Keonne Rodriguez and William Lonergan Hill on charges of conspiracy to commit money laundering and conspiracy to operate an unlicensed money-transmitting business. The DOJ filing alleged that Samourai had processed approximately 100 million in money laundering. The indictment seized the Samourai web infrastructure; the mobile applications were removed from Apple’s App Store and Google Play; the Whirlpool service ceased operating.
  • Aftermath: the indictment had a chilling effect on the broader Bitcoin-privacy-tool ecosystem. zkSNACKs cited the Samourai case in its decision to shut down the Wasabi coordinator in June 2024.

The Samourai case resolved in 2025 — both founders pled guilty to conspiracy to operate an unlicensed money-transmitting business (July 2025) and were sentenced in November 2025 — without litigating the central question it raised: the legal status of operating a non-custodial privacy tool (the Samourai team did not custody user funds; Whirlpool’s architecture is structurally similar to JoinMarket in terms of trust model). The guilty pleas left that question unresolved, and it remains closely watched by the broader privacy-tool community (see Tornado Cash sanctions and the privacy-tool regulatory landscape).

The 2024 wind-downs in operational terms

The April 2024 Samourai indictment and June 2024 Wasabi shutdown removed two major CoinJoin coordinators from operational service. As of 2026:

  • Wasabi software continues to exist but the canonical zkSNACKs coordinator is offline. Community-operated coordinators have emerged; the privacy properties depend heavily on the operator’s discipline and trust. Community maintenance has kept the software itself moving (v2.8.0, June 2026 — see above).
  • Samourai Whirlpool is offline; the seized infrastructure is unavailable; the Whirlpool architecture is documented but not in active operation.
  • JoinMarket continues to operate without disruption.
  • Sparrow Wallet’s CoinJoin integration continues with workarounds; depends on remaining coordinator infrastructure.
  • Newer implementations (BTCPay Server’s experimental CoinJoin features; community-operated coordinators) provide partial replacement.

The cumulative effect: CoinJoin remains technically available in 2026, but the operational landscape is materially worse than the 2018-2024 peak. The user base has consolidated to JoinMarket and Sparrow-with-community-coordinators; the wallet UX is rougher; the anonymity sets are smaller; the legal-uncertainty footprint is higher.


Operational reality in 2026

For a holder choosing to CoinJoin in 2026, the practical landscape:

Tool selection

JoinMarket — the sovereign tool. Run your own node; configure your IRC participation; participate as taker (paying fees to mix) or maker (earning fees by providing liquidity). Operationally complex; the steepest learning curve; the strongest sovereign properties; survived the 2024 enforcement actions intact.

Sparrow Wallet with community coordinators — friendlier UX; coordinator selection matters substantially for privacy properties; trust assumptions on the coordinator must be evaluated.

Wasabi with community coordinators — actively maintained again as a community open-source project (v2.8.0, June 2026: P2P compact-filter sync with no central indexing server, experimental Pay-in-CoinJoin, payment batching); community coordinators provide partial replacement; coordinator-trust evaluation remains the critical step.

BTCPay Server experimental CoinJoin — emerging; operationally rougher; for technical users willing to participate in early-stage tooling.

The honest assessment: none of these provides the ease-of-use of 2020-2024 Wasabi. CoinJoin in 2026 is meaningfully harder than CoinJoin in 2022; the usability regression is real.

Operational considerations

Mix denominations: most pools operate at specific denominations (0.01, 0.05, 0.1 BTC are common). The holder’s UTXO sizes need to fit the pool denominations; this often requires a pre-mix consolidation or splitting step.

Anonymity sets: smaller pools mean smaller anonymity sets. A pool with 100 participants over the past month is more private than a pool with 10 participants over the past month. Evaluate the actual recent transaction history of any pool before committing UTXOs.

Fees: coordinator fees (typically 0.3-1% in the Wasabi/Samourai era; lower for JoinMarket) plus mining fees. For modest mixing amounts the cost is not trivial.

Time commitment: rounds may take minutes to hours to complete depending on participant availability. For Whirlpool-style pre-mix architectures, time-to-mix could be hours or days for less-popular denominations.

Post-mix UTXO discipline: the load-bearing operational requirement. Post-mix UTXOs must be managed separately from pre-mix UTXOs to preserve clustering ambiguity. Labelled-wallet tooling (Sparrow, Specter) is essential.

Exchange-deposit compliance friction: many KYC exchanges will freeze deposits of post-CoinJoin coins pending compliance review. Some (Kraken, others) have policies that effectively prevent post-CoinJoin deposits. Plan the off-ramp before mixing.

When CoinJoin makes sense

For a holder whose threat model justifies it AND who can absorb the operational and compliance friction:

  • Post-KYC withdrawal hygiene: withdraw a portion from a KYC exchange to self-custody, CoinJoin, then use the post-mix outputs for receive activity going forward. Limits the propagation of KYC-attached clustering.
  • Substantial holdings privacy: for Tier 2+ holdings, raising the chain-analysis cost is a meaningful threat-model defence against the chain-analysis-to-physical-attack pipeline.
  • Specific privacy needs: holders with specific privacy threat models (journalists, activists, businesses with adversarial commercial counterparties) benefit from the explicit privacy practice.

For a holder without those conditions:

  • Address-reuse hygiene plus the tiered-wallet pattern captures much of the practical privacy benefit without CoinJoin’s operational and compliance friction.
  • Non-KYC acquisition channels (Bisq, Hodl Hodl) for the privacy-sensitive portion of the stack can substitute for post-KYC CoinJoin in many cases.
  • Lightning use for transactional activity (see Lightning privacy properties) provides off-chain privacy that does not carry the CoinJoin compliance footprint.

What changes if you do CoinJoin

The cluster of post-mix UTXOs is clustering-ambiguous from the mix forward. Forward spending propagates that ambiguity to recipients (if discipline is maintained). The pre-mix cluster retains its prior identity-resolution; the chain-analysis firm sees a clear “this person CoinJoined” event but cannot trace post-mix outputs back to the holder with certainty.

What you trade: time, fees, operational complexity, KYC-exchange-deposit friction, the on-chain footprint that flags your wallet as a CoinJoin participant, and (post-2024) some legal-uncertainty exposure that did not exist as starkly before.


Counter-arguments and tensions

”CoinJoin anonymity-set claims are overstated”

The argument: Wasabi marketed “100-participant anonymity sets” but the structural anonymity set depends on Sybil resistance, which is hard to verify. Empirical research suggests effective anonymity sets are smaller than marketed.

Response: Substantive critique. Several academic papers (Möser et al. 2018; Maurer et al. 2019; subsequent work) demonstrated that practical anonymity sets in early Wasabi were materially smaller than the apparent participant counts due to wallet-fingerprint leakage and structural correlations. Wabisabi’s 2022 redesign addressed some of these issues; the 2024 wind-down halted further iteration. The defensible position: CoinJoin provides meaningful clustering-ambiguity but the anonymity-set numbers should be discounted, particularly for older implementations. A holder evaluating CoinJoin should expect “post-mix outputs are clustering-ambiguous to ~10-30 effective participants” rather than the marketed numbers, and design their post-mix discipline accordingly.

”CoinJoin participation is itself a flag”

The argument: Chain-analysis firms identify CoinJoin transactions trivially. The CoinJoin participant signals to chain-analysis “this wallet is doing privacy stuff,” which can prompt enhanced compliance review at exchanges or other downstream KYC venues. The privacy-practice is self-defeating because it draws attention.

Response: Real tension. The flag exists; the compliance consequences are real (some exchanges have frozen accounts that deposited post-CoinJoin coins). Two honest responses: (1) the flag matters most for holders whose threat model is dominated by regulatory exposure; for holders whose threat model is dominated by physical-attack targeting, the flag is a cost worth paying; (2) the compliance landscape is asymmetric — some exchanges are aggressive flaggers, others are not — and the holder can structure their off-ramp choice around this asymmetry. The right answer depends on the holder’s specific situation. See Threat modeling for self-custody for the framework.

”The Samourai indictment chilled the developer ecosystem; CoinJoin is a dying technology”

The argument: The 2024 enforcement actions removed two of three major coordinators, chilled new development, and signal that the regulatory direction of travel is hostile. Investing in CoinJoin practice in 2026 is investing in a deprecated technology.

Response: Partially true; the trajectory is genuinely uncertain. JoinMarket survives because of its sovereign architecture; Sparrow’s CoinJoin integration continues; community-operated coordinators have partial substitution. But the development pace and user-experience have regressed materially. The honest read: CoinJoin in 2026 is a meaningfully degraded version of CoinJoin in 2022, and the trajectory is not clearly improving. For holders prioritising privacy practice in 2026, CoinJoin remains available but the realistic horizon for further capability improvement is uncertain. Adjacent tools — PayJoin, Stealth addresses and Silent Payments, Lightning privacy properties — represent complementary or alternative paths. The Privacy practice cluster doesn’t bet everything on CoinJoin specifically.

”Operating a non-custodial CoinJoin tool should be First-Amendment-protected speech”

The argument: The Samourai team did not custody user funds; their software is code. Indicting them for users’ downstream conduct is equivalent to indicting cryptography developers for what users do with cryptography. The case raises substantial First Amendment and chilling-effect concerns.

Response: The legal-theory critique has merit and is being argued in the ongoing Samourai proceedings. The outcome will materially shape the legal landscape for privacy-tool development. From the practical-holder perspective, the relevant points: (1) regardless of how the case resolves, the chilling effect on current development is real; (2) JoinMarket’s sovereign architecture is partially insulated from this particular legal theory because there is no operator to indict; (3) the regulatory direction more broadly is hostile to non-custodial privacy tools, and the holder should plan their privacy stack on the assumption that the landscape may worsen further before it improves. This is a “real but unsettled” situation; the holder should track developments but not stake their privacy practice on optimistic legal outcomes.

”CoinJoin is overkill for most holders; address-reuse hygiene plus tiered wallets is enough”

The argument: The privacy benefit from CoinJoin is marginal compared to the privacy benefit from basic address-reuse hygiene. Most holders should focus on the basics and not adopt the operational complexity of CoinJoin.

Response: Largely true for the median holder. Address-reuse hygiene plus the tiered-wallet pattern captures most of the practical privacy benefit available at low operational cost. CoinJoin is the next layer up; it makes sense for holders whose threat model justifies the additional friction. The Privacy practice cluster’s structure reflects this: Address reuse and chain analysis is the foundational threat note, hygiene is the foundational defence, and CoinJoin is one specific tool for holders ready to go further. The holder shouldn’t feel obliged to CoinJoin; the holder should feel obliged to hygiene.


Open questions for further development

  • The Samourai case’s legal outcome will materially shape the privacy-tool landscape. How should the section handle the legal-uncertainty period? Tracking active updates is operationally expensive.
  • JoinMarket’s user base is small relative to the historic Wasabi/Samourai base. Is JoinMarket’s anonymity-set sufficient for practical privacy in 2026, or has it become too small to provide the protection it once did?
  • The compliance footprint at KYC exchanges varies widely. A jurisdictional map of “which exchanges flag CoinJoin deposits and which don’t” would be useful operational content; this changes faster than the section can sustainably track.
  • The community-coordinator landscape for Wasabi software post-2024 is in flux. Which operators have demonstrated reliable privacy properties and are likely to persist?
  • Newer privacy primitives (Stealth addresses and Silent Payments, emerging BIP proposals) may substitute for CoinJoin functionality in some use cases. What’s the realistic timeline for that substitution, and what should the section say about the current allocation of effort between CoinJoin and adjacent tools?

Canonical sources for this note

Foundational:

  • Maxwell, Gregory — CoinJoin: Bitcoin privacy for the real world (bitcointalk.org post, August 22, 2013) — the original proposal
  • See Greg Maxwell for the thinker treatment

JoinMarket:

  • Adam Gibson (waxwing) — JoinMarket documentation and design papers
  • JoinMarket-org GitHub — implementation source, user documentation
  • Mastering Bitcoin (Antonopoulos) — covers JoinMarket in privacy chapters

Wasabi:

  • zkSNACKs technical papers on the Chaumian CoinJoin protocol
  • WabiSabi paper (2021) — credential-based CoinJoin upgrade
  • Wasabi documentation (archived post-2024)
  • zkSNACKs June 2024 shutdown announcement
  • Wasabi Wallet v2.8.0 release notes (GitHub, June 27, 2026) — the community-maintained project’s post-shutdown feature release

Samourai:

  • Samourai Wallet documentation (much archived post-2024 seizure)
  • Whirlpool whitepaper and protocol specification
  • US v. Rodriguez et al., S.D.N.Y. (indictment April 24, 2024) — the DOJ filing
  • Coverage in CoinDesk, Bitcoin Magazine, The Block (2024)

Academic analysis:

  • Möser, Soska, Heilman, Lee, Heffan, et al. — An Empirical Analysis of Linkability in the Monero Blockchain and related Bitcoin CoinJoin analyses (2018-2022)
  • Maurer, Neudecker, et al. — Anonymous CoinJoin Transactions with Arbitrary Values (2017)
  • Subsequent academic work on Wasabi anonymity-set empirical measurement

Practitioner literature:

  • 6102bitcoin’s Bitcoin Privacy Guide — operational walkthrough including CoinJoin
  • Bitcoin Q+A podcast — episodes covering Wasabi/Samourai/JoinMarket comparisons
  • Sparrow Wallet documentation — CoinJoin operational notes

Legal-and-regulatory context:

  • Coin Center analyses of CoinJoin legal status (pre- and post-Samourai)
  • Electronic Frontier Foundation commentary on privacy-tool development chilling effects
  • Coverage in the broader crypto-policy press (Coin Center, Bitcoin Policy Institute, etc.)

As of 2026: the Samourai case resolved in guilty pleas (July 2025) and sentencing (November 2025); the JoinMarket and community-coordinator landscape continues to evolve; Wasabi software development is active again under community maintenance (v2.8.0, June 2026) though the canonical zkSNACKs coordinator remains offline.


Companion Privacy practice notes:

The threat-modeling framework:

Operational practice:

Custody configurations:

Adjacent thinker pages:

  • Greg Maxwell — original CoinJoin proposer; cryptographic thinker
  • Jameson Lopp — practitioner-perspective on privacy practice

The sub-MOC home: