Skip to main content

Boteraser | Website and Server Security Solutions

🛡️ CVE-2026-35040 — fast-jwt

🟡 CVSS 5.3 — Medium ⚠️ Exploit Public CWE-440 OSV
5.3
CVSS Score
0 Low4 Medium7 High9 Critical10

Description

fast-jwt: Stateful RegExp (/g or /y) causes non-deterministic allowed-claim validation (logical DoS)

Impact

Using certain modifiers on RegExp objects in the allowedAud, allowedIss, allowedSub, allowedJti, or allowedNonce options in verify functions can cause certain unintended behaviours. This is because some modifiers are stateful and will cause failures in every second verification attempt regardless of the validity of the token provided.

Such modifiers are:

  • /g : Global matching
  • /y : Sticky matching

This does NOT allow invalid tokens to be accepted, only for valid tokens to be improperly rejected in some configurations. Instead it causes 50% of valid authentication requests to fail in an alternating pattern, leading to:

  • Intermittent user authentication failures
  • Potential retry storms in applications
  • Operational monitoring alerts

Affected Configurations

This vulnerability ONLY affects applications that:

  • Use RegExp objects (not strings) in the allowedAud, allowedIss, allowedSub, allowedJti, or allowedNonce options
  • Use stateful RegExp modifiers such a /g or /y

Example: allowedAud: /abc/g ← IMPACTED

Example: allowedAud: "/abc/" ← SAFE

Not Affected

  • Applications using string patterns for audience validation (most common)
  • Applications using RegExp patterns without stateful modifiers

Assessment Guide

To determine if you're affected:

Check if allowedAud, allowedIss, allowedSub, allowedJti, or allowedNonce options use RegExp objects (/pattern/ or new RegExp())

If yes, review the pattern for stateful modifiers like /g, /y

If no RegExp usage or no stateful modifiers, you are NOT affected

Mitigation Options

While a fix will be coming in the next version of the package you can take steps to mitigate the issue immediately by removing any such modifiers (/g, /y) from the regex.

Summary

fast-jwt accepts RegExp for allowedAud, allowedIss, allowedSub, allowedJti, and allowedNonce.

If the provided regular expression uses the g (global) or y (sticky) flag, verification becomes non-deterministic: the same valid token alternates between acceptance and rejection across successive calls.

This occurs because RegExp.prototype.test() is stateful when g/y is set (it mutates lastIndex), and fast-jwt reuses the same RegExp object without resetting lastIndex.

Affected component

src/verifier.js

ensureStringClaimMatcher() returns the RegExp object directly.

validateClaimValues() performs repeated a.test(v) calls without resetting lastIndex.

Impact

Logical denial-of-service / authentication flapping.

A valid signed JWT can be intermittently rejected.

Causes unpredictable authentication outcomes across repeated verification calls.

Can trigger retry storms and cascading failures in API gateways and authentication middleware.

Affects any deployment that configures allowed* using RegExp and includes g or y flags.

Root cause

validateClaimValues() uses: allowed.some(a => a.test(v))

When a is a RegExp with g or y, a.test() mutates a.lastIndex.

Subsequent calls against the same input can return different results.

Proof of concept

Environment

  • fast-jwt: 6.1.0 (repo HEAD)
  • Node.js: v24.13.1

PoC

const { createSigner, createVerifier } = require('fast-jwt')

const sign = createSigner({ key: 'secret' })

const token = sign({ aud: 'admin', iss: 'issuer' })

function run(name, opts) {

const verify = createVerifier({ key: 'secret', ...opts })

console.log('\n==', name)

for (let i = 0; i < 8; i++) {

try { verify(token); console.log(i, 'PASS') }

catch (e) { console.log(i, 'FAIL', e.code || e.message) }

}

}

run('allowedAud global regex', { allowedAud: /^admin$/g })

run('allowedIss global regex', { allowedIss: /^issuer$/g })

run('control (non-global regex)', { allowedAud: /^admin$/ })

Observed behavior

  • allowedAud with /g alternates PASS/FAIL across calls
  • allowedIss with /g alternates PASS/FAIL across calls
  • control regex (no g/y) is deterministic and always PASS

Expected behavior

Validation must be deterministic.

The same token under the same verifier configuration must always yield the same decision.

Suggested fix (minimal and safe)

Wrap RegExp matchers inside ensureStringClaimMatcher() to reset lastIndex before calling test():

if (r instanceof RegExp) {

return { test: v => { r.lastIndex = 0; return r.test(v) } }

}

This preserves semantics for non-global regexes, makes g/y deterministic, and avoids changes in the rest of the verifier logic.

Security classification

Logical DoS / authentication reliability failure.

This can be weaponized to produce production outages via retry storms and auth instability.

Why this is not “misuse”

  • The library explicitly accepts RegExp for allowed* claim validation.
  • The behavior difference is caused by internal state mutation of RegExp.test().
  • The same token, same verifier config, same runtime yields different outcomes.
  • Security decisions must be deterministic; non-determinism at the verification layer is a co

How this vulnerability can be exploited

This issue can be reached over the network, attack complexity is low, an attacker needs no privileges on the target. No user interaction is required. The scope is unchanged, so the impact stays within the vulnerable component. Rated impact: confidentiality none, integrity none, availability low.

CVSS metrics in full

The score comes from this vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L

  • Attack vector: Network — reachable from anywhere that can route to the service.
  • Attack complexity: Low — the attack works reliably, with no preparation.
  • Privileges required: None — an unauthenticated stranger can try it.
  • User interaction: None — nobody has to be tricked into anything.
  • Scope: Unchanged — the damage stays inside the vulnerable component.
  • Confidentiality impact: None.
  • Integrity impact: None.
  • Availability impact: Low — limited, and the attacker does not choose what is affected.

Weakness class

CVE-2026-35040 is classified as CWE-440: Expected Behavior Violation. A feature, API, or function does not perform according to its specification.

Affected software

CVE-2026-35040 is recorded against 1 package.

  • fast-jwt

Timeline and source

Published on 9 April 2026 and last revised on 17 June 2026. A public exploit is known to exist, which raises the urgency of patching considerably. A vendor advisory or fix has been published. Record sourced from OSV.

References

github.com (Web)
nvd.nist.gov (Advisory)
github.com (Web)
github.com (Web)
github.com (Package)
github.com (Web)

Other advisories for this package

fast-jwt has other advisories on record. If you are patching this one, these are worth checking on the same host:

Same weakness in other software

These advisories are the same class of weakness (CWE-440: Expected Behavior Violation) in other software:

Details

Severity MEDIUM
CVSS Score 5.3
CVSS Vector CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L
CWE CWE-440
Public Exploit ⚠️ Yes
Source OSV
Published 2026-04-09
Updated 2026-08-20
Modified 2026-06-17

Affected Packages

Software From version Fixed in
fast-jwt

Similar Threats

Exploit Protection

Are you running fast-jwt?

CVE-2026-35040 carries CVSS 5.3 Medium rating and a public exploit already exists. BotEraser checks your installation against this and other known CVE records, and blocks IPs associated with exploit activity.

Check My Site For CVE-2026-35040 →

No credit card required  ·  Results in minutes

ⓘ Data Notice: The information presented above has been compiled from publicly available internet sources. Boteraser aggregates this data solely for informational purposes and does not independently classify, evaluate, or endorse any findings about the vulnerabilities listed. The accuracy and completeness of this information is the sole responsibility of the original publishers. Boteraser and its operators accept no liability for any decisions made based on this data.

Browse related advisories

All advisoriesCVECVE 2026