PipeFlare

What Is a Bitcoin Covenant?

A Bitcoin covenant would restrict where an output can be spent next. Compare proposed transaction templates, stack operations, vaults and their current status.

By Jonathan West · Updated November 17, 2026

A Bitcoin covenant is a proposed rule in an output's locking script that would restrict where the coins can be spent next

Covenant proposals describe different ways to limit future spends, but none of the proposals compared here changes how Bitcoin outputs work today

Category

Protocol proposal

Difficulty

Intermediate

Where you'll see it

Bitcoin Improvement Proposal discussions about transaction templates, tapscript, vaults and future consensus rules

First introduced

BIP-119 (CTV) was assigned in January 2020, and Bitcoin Optech lists covenant discussions from 2018 onward

About bitcoin covenants

A Bitcoin covenant is a proposed rule in a coin's locking script that restricts where the coin can be spent next. Four proposals describe different ways to support these restrictions through soft forks: OP_CHECKTEMPLATEVERIFY (CTV) in Bitcoin Improvement Proposal 119 (BIP-119), OP_CAT in BIP-347, OP_VAULT in BIP-345 and OP_CHECKCONTRACTVERIFY in BIP-443. A soft fork changes consensus rules. None of these four proposals has a deployed status.

What a Bitcoin Covenant Is

A Bitcoin covenant would let an output's locking script restrict the locking scripts of outputs created when that coin is spent. Bitcoin Optech quotes Anthony Towns's definition: "In the context of Bitcoin, the most useful definition of covenant is that it's when the scriptPubKey of a UTXO restricts the scriptPubKey in the output(s) of a tx spending that UTXO." An unspent transaction output (UTXO) is a coin available to spend, and its scriptPubKey is the locking script that sets its spending conditions.

The distinction is about the next output. Optech gives an example in which a covenant permits a spend only to approved scripts, such as a script returning coins to the same user's balance or a staging output with a delay before further spending. A spender could meet the conditions for spending the original output and still fail the covenant's restriction on the destination output.

Optech describes covenants as proposed changes to Bitcoin's consensus rules. The four proposals compared here do not provide a general covenant feature on the Bitcoin network today. Adding a consensus restriction would require a Bitcoin soft fork, and the details depend on which proposal is under discussion.

Why People Want Covenants

Covenant proposals address specific limits on how an output can be spent next. BIP-119 describes a transaction template that commits to selected details of a later spending transaction. Its stated aim includes reducing the trust, interaction and storage demands of applications that otherwise rely on transactions signed in advance.

BIP-345 describes a vault with a delayed withdrawal and a separate recovery path. Its authors present that arrangement as a way to intervene after a spending key is compromised, before a withdrawal is final. Bitcoin Optech's Ark topic says Ark can be implemented without a consensus change, while covenant features such as CTV could support more users and improve fee efficiency. Those are proposed construction benefits, not an activated feature.

Bitcoin already has scripts that set spending conditions. Taproot's BIP-341 describes script trees, while BIP-347 identifies a missing general-purpose way to combine stack values in tapscript. Whether existing Bitcoin scripts count as smart contracts depends on the definition. A general rule restricting the scripts of future outputs is the covenant capability these proposals discuss.

The Main Proposals Side by Side

The four proposals specify different restrictions or building blocks, so their names are not interchangeable. CTV commits to a transaction template, OP_CAT combines stack values, OP_VAULT describes a delayed withdrawal with recovery, and OP_CHECKCONTRACTVERIFY (OP_CCV) describes output restrictions that can carry committed data. Each BIP identifies its proposed consensus change as a soft fork.

Compared against the BIP texts as read on 17 November 2026. "Complete" in the OP_CAT header describes the specification's status. It does not mean the opcode is activated. Bitcoin Optech lists many more covenant proposals and related discussions than the four shown here.

Covenant proposals at a glance
BIPNameStatus at retrievalWhat it doesSource
BIP-119OP_CHECKTEMPLATEVERIFYDraftChecks a spending transaction against a committed template hash. Proposed as a soft fork.BIP-119 text
BIP-347OP_CATCompleteJoins two tapscript stack values. Proposed as a soft fork.BIP-347 text
BIP-345OP_VAULTClosed, with proposed replacement BIP-443Describes a delayed withdrawal and recovery path using proposed vault opcodes with CTV.BIP-345 text
BIP-443OP_CHECKCONTRACTVERIFYDraftProposes output restrictions and UTXOs carrying a dynamic commitment to data.BIP-443 text

How it actually works

A covenant changes the question a locking script can answer when a transaction spends an output. The script would check a condition on the resulting output or transaction, rather than only conditions for authorizing the current spend. Bitcoin Optech's example allows a spend to a staging output but limits what that output's script permits afterward.

The proposed checks differ. Under BIP-119, a script would compare a hash supplied on its stack with a hash calculated from specified fields of the spending transaction. Under BIP-347, OP_CAT would join two stack values. Other script operations could then use the combined value in constructions described by the BIP. BIP-345 instead describes dedicated vault and recovery opcodes used alongside CTV, while BIP-443 proposes a more general way to constrain output details and carry committed data.

Each mechanism would need its proposed consensus rule before Bitcoin nodes could enforce it. A BIP documents the proposed rule and its status. A written specification alone does not change the network. That difference matters when a diagram shows how a vault or template would work: it can explain a proposed transaction path without showing a feature available for ordinary outputs today.

How CTV Works

CTV in BIP-119 would check whether a spending transaction matches a hash committed in the script. The proposed OP_CHECKTEMPLATEVERIFY opcode gives the existing OP_NOP4 opcode a new meaning after activation. Before activation, BIP-119 specifies that it continues to act as OP_NOP4.

The proposed check takes a value from the script stack. An empty stack fails the check. A top value of 32 bytes is compared with the transaction's calculated template hash. A mismatch fails. The hash commits to transaction fields including the version, locktime, input sequences, output count, outputs and the index of the input being checked.

A template therefore describes permitted transaction details in advance. It does not mean that every later transaction must have the same recipients: the restriction would apply to an output whose script uses the proposed opcode and a particular template hash. The BIP-119 specification lists the exact committed fields for readers comparing CTV with proposals that constrain outputs in other ways.

How OP_CAT Works

OP_CAT in BIP-347 would join the top two values on a tapscript stack into one value. With stack values x1 and x2, where x2 is on top, the result is x1 followed by x2. The operation fails if there are fewer than two values or if the combined value exceeds the 520-byte script-element limit.

BIP-347 proposes redefining the tapscript opcode OP_SUCCESS126 as OP_CAT through a soft fork. The proposed change is scoped to tapscript. The BIP says use of OP_CAT outside tapscript continues to trigger a disabled-opcode error. Its "Complete" header status refers to the BIP specification, not network deployment.

Unlike CTV, OP_CAT is an operation for assembling data inside a script rather than a direct check against a fixed transaction template. BIP-347 lists potential constructions including tree signatures, vaults and arbitrary computation on larger stack elements. Bitcoin Optech also links a discussion of when enabling OP_CAT could allow recursive covenants. That possibility is part of the proposal debate, not an active network rule.

How OP_VAULT Works and Where It Stands

OP_VAULT in BIP-345 describes a specialized covenant for delayed withdrawals with a recovery route. The proposal combines OP_VAULT and OP_VAULT_RECOVER with CTV. A withdrawal to an arbitrary destination would have to pass through a delay, while a prespecified recovery path could be used before the withdrawal becomes final.

The delay gives the recovery path a specific job in the proposal. BIP-345 describes a case in which a spending key is compromised and the coin's custodian notices an unexpected withdrawal attempt. The recovery path would permit an intervening spend before that withdrawal completes, provided the proposed rules were activated and the output had been set up to use them.

BIP-345's header says "Closed" and names BIP-443 as its proposed replacement. BIP-443 is a Draft for OP_CHECKCONTRACTVERIFY, a broader proposed primitive that can constrain an output's internal public key and taptree and carry a commitment to data. A closed vault BIP and a draft replacement should not be read as an available vault feature.

Where Covenants Would Fit After Taproot

A Bitcoin covenant opcode would need a consensus change before full nodes could enforce it. Taproot's BIP-341 has a Deployed header status, which makes it a useful contrast with the Draft, Complete and Closed statuses in this comparison. The site's background on Taproot and SegWit explained covers the script system these proposals would extend.

Each of the four BIPs describes a soft-fork proposal, but their proposed rules differ. A soft fork would change what a node enforces when checking affected transactions. The Bitcoin soft fork explainer covers that consensus mechanism. A BIP header's status must be read separately from the mechanism it describes.

The possible applications differ too. Optech says Ark does not require a consensus change to be implemented, although features like CTV could alter its capacity and fee efficiency. BIP-443 also names Ark among constructions that could use output restrictions. Neither statement makes Ark dependent on activation of any one covenant proposal.

What Covenants Mean for Your Own Coins

The four covenant proposals compared here do not change the spending rules for Bitcoin outputs today. None has a BIP header status of Deployed in the documents read for this comparison. BIP-345's delayed vault withdrawal is a description of proposed consensus behavior, not a setting that can be enabled for an ordinary output under that proposal.

If a proposed covenant rule were activated, its restriction would apply to outputs constructed to use it. The proposals describe new tapscript opcodes or, in CTV's case, a new meaning for OP_NOP4. An output that did not use the new rule would not acquire a vault delay or transaction template merely because another output used one.

For a reader examining a proposed spend, the useful question is which output's script contains the restriction. With CTV, identify the committed template hash and the transaction fields it covers. With a vault design, identify the withdrawal delay and the prespecified recovery path in the proposal's description.

Takeaway on Bitcoin Covenants

Bitcoin covenants are a family of proposed spending restrictions, with different BIPs describing different ways to constrain a later transaction or output. CTV specifies a template check. OP_CAT supplies a stack operation that could support covenant constructions. The closed OP_VAULT proposal describes delayed withdrawal with recovery. The draft OP_CHECKCONTRACTVERIFY proposal describes broader output restrictions.

These distinctions apply to the BIP statuses and specifications read on 17 November 2026. A change to a proposal's specification, header status or activation would require a fresh comparison. To check a Bitcoin covenant proposal directly, read its status header and specification in the official BIP repository.

Start here

  1. 1Read [Bitcoin Optech's covenant definition](https://bitcoinops.org/en/topics/covenants/) and identify the original output and the later output it describes.
  2. 2Open the [BIP-119 header and specification](https://github.com/bitcoin/bips/blob/master/bip-0119.mediawiki). Locate its status and the fields committed by the template hash.
  3. 3Compare the proposed stack operation in [BIP-347](https://github.com/bitcoin/bips/blob/master/bip-0347.mediawiki) with the dedicated delayed-withdrawal path in [BIP-345](https://github.com/bitcoin/bips/blob/master/bip-0345.mediawiki).
  4. 4Read the [BIP-345 header](https://github.com/bitcoin/bips/blob/master/bip-0345.mediawiki) for its proposed replacement, then open [BIP-443](https://github.com/bitcoin/bips/blob/master/bip-0443.mediawiki) to see how its output restrictions differ.
  5. 5Use [BIP-341's header](https://github.com/bitcoin/bips/blob/master/bip-0341.mediawiki) as an example of a Deployed status when interpreting the statuses of covenant proposals.

Strengths

  • CTV's proposed template check specifies transaction fields that an output could require in its later spend.
  • BIP-345 describes a recovery route that could intervene during a delayed vault withdrawal.
  • OP_CAT and OP_CHECKCONTRACTVERIFY describe different building blocks for output restrictions, giving readers concrete mechanisms to compare.

Common misunderstandings

  • A BIP header marked Complete does not mean its proposed opcode is active. BIP-347 is a specification with no Deployed status.
  • OP_CAT joins stack values. Its proposed operation is different from CTV's direct comparison with a transaction template hash.
  • A closed BIP is not an activated feature. BIP-345 names a proposed replacement in its header.
  • Ark can be implemented without a covenant consensus change, according to Bitcoin Optech, although covenant features could support more users and greater fee efficiency.

Common questions

What is a Bitcoin soft fork?

A Bitcoin soft fork changes Bitcoin's consensus rules. A covenant opcode would need such a rule change before nodes could enforce its proposed restrictions. See what a Bitcoin soft fork is for the consensus mechanics.

Are there smart contracts on Bitcoin?

Bitcoin outputs already use scripts to set spending conditions, and Taproot's BIP-341 added script-tree capabilities. Whether those conditions are called smart contracts depends on the definition. The general ability to restrict the locking scripts of later outputs is the proposed covenant feature discussed here.

What is OP_CTV?

OP_CTV is a shorthand name for the proposed OP_CHECKTEMPLATEVERIFY opcode in BIP-119. It would compare a hash committed by an output's script with a hash of specified fields in the transaction spending that output. BIP-119 is a Draft, so its template check is not an activated consensus rule.

What is OP_CAT?

OP_CAT is the proposed tapscript operation in BIP-347 that joins the top two stack values. The combined value cannot exceed 520 bytes. BIP-347's Complete status describes its specification, not deployment of the opcode.

What is a Bitcoin vault?

A Bitcoin vault, as described in BIP-345, is a proposed arrangement that delays a withdrawal to an arbitrary destination while allowing a prespecified recovery path before final withdrawal. BIP-345 is Closed and names BIP-443 as its proposed replacement. The BIP's vault behavior is not an activated Bitcoin feature.

Is this financial advice?

No. This page explains proposed Bitcoin script rules and BIP statuses for general information. It is not financial advice.

Sources

Related guides

Ready to put this into practice?

Exchange sign-up bonuses pay both you and a referrer after a qualifying trade.

See bonuses →