Bump the npm_and_yarn group across 4 directories with 7 updates

[dependabot]

Bumps the npm_and_yarn group with 4 updates in the / directory: undici, esbuild, follow-redirects and form-data.
Bumps the npm_and_yarn group with 5 updates in the /packages/plugin-hardhat/examples/BoxSolidityTests directory:

Package	From	To
undici	6.26.0	6.27.0
undici	8.3.0	8.5.0
esbuild	0.28.0	0.28.1
form-data	4.0.5	4.0.6
js-cookie	2.2.1	3.0.8
ws	8.20.1	8.21.0

Bumps the npm_and_yarn group with 5 updates in the /packages/plugin-hardhat/examples/BoxTransparent directory:

Package	From	To
undici	6.26.0	6.27.0
undici	8.3.0	8.5.0
esbuild	0.28.0	0.28.1
form-data	4.0.5	4.0.6
js-cookie	2.2.1	3.0.8
ws	8.17.1	8.21.0

Bumps the npm_and_yarn group with 5 updates in the /packages/plugin-hardhat/examples/BoxUUPS directory:

Package	From	To
undici	6.26.0	6.27.0
undici	8.3.0	8.5.0
esbuild	0.28.0	0.28.1
form-data	4.0.5	4.0.6
js-cookie	2.2.1	3.0.8
ws	8.17.1	8.21.0

Updates undici from 8.3.0 to 8.5.0

Release notes

Sourced from undici's releases.

v8.5.0

⚠️ Security Release

This release line addresses 8 security advisories. Most are fixed in v8.5.0; the SOCKS5 pool-reuse issue was fixed earlier in v8.2.0.

Action required: Upgrade to undici 8.5.0 or later.

npm install undici@^8.5.0

Summary

Advisory	CVE	Severity (CVSS)	Fixed in	Fix commit
GHSA-vxpw-j846-p89q	CVE-2026-12151	High (7.5)	8.5.0	32dbf0b3
GHSA-38rv-x7px-6hhq	CVE-2026-9675	High (7.5)	8.5.0	b4c287b3
GHSA-vmh5-mc38-953g	CVE-2026-9697	High (7.4)	8.5.0	42d49559
GHSA-hm92-r4w5-c3mj	CVE-2026-6734	High (7.5)	8.2.0	a516f870
GHSA-pr7r-676h-xcf6	CVE-2026-9678	Moderate (5.9)	8.5.0	cb105d7c
GHSA-p88m-4jfj-68fv	CVE-2026-9679	Moderate (5.9)	8.5.0	5655ea43
GHSA-g8m3-5g58-fq7m	CVE-2026-11525	Low (3.7)	8.5.0	5655ea43
GHSA-35p6-xmwp-9g52	CVE-2026-6733	Low (3.7)	8.5.0	6ea54ef8

---

High severity

WebSocket DoS via fragment count bypass — CVE-2026-12151

GHSA-vxpw-j846-p89q · CWE-400, CWE-770 Fix: 32dbf0b3 websocket: limit the number of fragments in a message (also c5ed7875 handle empty fragments and stream limits)

A malicious WebSocket server can stream a large number of small or empty continuation frames. Undici enforced a limit on cumulative payload size but did not limit the number of fragments per message, leading to unbounded memory growth and denial of service.

• Affected: applications using new WebSocket(...) or WebSocketStream against untrusted endpoints.
• Workaround: none — upgrade is required.

WebSocket DoS via cumulative fragment bypass — CVE-2026-9675

GHSA-38rv-x7px-6hhq · CWE-400, CWE-770 Fix: b4c287b3 fix(websocket): enforce max payload size across fragments

Undici validated the size of individual frames but did not track cumulative size across a fragmented message. An attacker could send many small fragments that each pass per-frame validation but collectively exceed the configured limit, causing memory exhaustion. This is a regression introduced in 8.1.0 (the

... (truncated)

Commits

• a0806e1 Bumped v8.5.0 (#5429)
• 8a0392c test: detect available python command in wpt runner (#5427)
• f4045b9 ci: increase Node.js workflow timeout (#5426)
• 363e44f chore: removed repro-h2-pipelining-default.mjs and lint (#5420)
• c5ed787 websocket: handle empty fragments and stream limits
• e114e77 align EventSource with spec (#5418)
• 6df53c5 fix: preserve h2 queue on out-of-order completion (#5410)
• 32dbf0b websocket: limit the number of fragments in a message
• 0d6ecc5 add bodymixin.textStream() (#5416)
• 42d4955 fix: honor requestTls when proxy is SOCKS5
• Additional commits viewable in compare view

Updates esbuild from 0.25.11 to 0.25.12

Release notes

Sourced from esbuild's releases.

v0.25.12

• Fix a minification regression with CSS media queries (#4315)

  The previous release introduced support for parsing media queries which unintentionally introduced a regression with the removal of duplicate media rules during minification. Specifically the grammar for @media <media-type> and <media-condition-without-or> { ... } was missing an equality check for the <media-condition-without-or> part, so rules with different suffix clauses in this position would incorrectly compare equal and be deduplicated. This release fixes the regression.

• Update the list of known JavaScript globals (#4310)

  This release updates esbuild's internal list of known JavaScript globals. These are globals that are known to not have side-effects when the property is accessed. For example, accessing the global Array property is considered to be side-effect free but accessing the global scrollY property can trigger a layout, which is a side-effect. This is used by esbuild's tree-shaking to safely remove unused code that is known to be side-effect free. This update adds the following global properties:

  From ES2017:

  • Atomics
  • SharedArrayBuffer

  From ES2020:

  • BigInt64Array
  • BigUint64Array

  From ES2021:

  • FinalizationRegistry
  • WeakRef

  From ES2025:

  • Float16Array
  • Iterator

  Note that this does not indicate that constructing any of these objects is side-effect free, just that accessing the identifier is side-effect free. For example, this now allows esbuild to tree-shake classes that extend from Iterator:

  // This can now be tree-shaken by esbuild:
  class ExampleIterator extends Iterator {}

• Add support for the new @view-transition CSS rule (#4313)

  With this release, esbuild now has improved support for pretty-printing and minifying the new @view-transition rule (which esbuild was previously unaware of):

  /* Original code */
  @view-transition {
    navigation: auto;
    types: check;
  }
  /* Old output */
  @​view-transition { navigation: auto; types: check; }
  /* New output */
  @​view-transition {
  navigation: auto;
  types: check;

... (truncated)

Changelog

Sourced from esbuild's changelog.

0.25.12

• Fix a minification regression with CSS media queries (#4315)

  The previous release introduced support for parsing media queries which unintentionally introduced a regression with the removal of duplicate media rules during minification. Specifically the grammar for @media <media-type> and <media-condition-without-or> { ... } was missing an equality check for the <media-condition-without-or> part, so rules with different suffix clauses in this position would incorrectly compare equal and be deduplicated. This release fixes the regression.

• Update the list of known JavaScript globals (#4310)

  This release updates esbuild's internal list of known JavaScript globals. These are globals that are known to not have side-effects when the property is accessed. For example, accessing the global Array property is considered to be side-effect free but accessing the global scrollY property can trigger a layout, which is a side-effect. This is used by esbuild's tree-shaking to safely remove unused code that is known to be side-effect free. This update adds the following global properties:

  From ES2017:

  • Atomics
  • SharedArrayBuffer

  From ES2020:

  • BigInt64Array
  • BigUint64Array

  From ES2021:

  • FinalizationRegistry
  • WeakRef

  From ES2025:

  • Float16Array
  • Iterator

  Note that this does not indicate that constructing any of these objects is side-effect free, just that accessing the identifier is side-effect free. For example, this now allows esbuild to tree-shake classes that extend from Iterator:

  // This can now be tree-shaken by esbuild:
  class ExampleIterator extends Iterator {}

• Add support for the new @view-transition CSS rule (#4313)

  With this release, esbuild now has improved support for pretty-printing and minifying the new @view-transition rule (which esbuild was previously unaware of):

  /* Original code */
  @view-transition {
    navigation: auto;
    types: check;
  }
  /* Old output */
  @​view-transition { navigation: auto; types: check; }
  /* New output */
  @​view-transition {
  navigation: auto;

... (truncated)

Commits

• 208f539 publish 0.25.12 to npm
• 5f03afd update release notes
• 6b2ee78 minify: remove css rules containing empty :is()
• f361deb add some additional known static methods
• 07aa646 automatically mark "RegExp.escape()" calls as pure
• 9039c46 simplify some call expression checks
• 188944d add some additional known static methods
• d3c67f9 fix #4310: add Iterator and other known globals
• 4a51f0b fix: escape dev server breadcrumb hrefs properly (#4316)
• 26b29ed fix #4315: @media deduplication bug edge case
• Additional commits viewable in compare view

Updates follow-redirects from 1.15.11 to 1.16.0

Commits

• 0c23a22 Release version 1.16.0 of the npm package.
• 844c4d3 Add sensitiveHeaders option.
• 5e8b8d0 ci: add Node.js 24.x to the CI matrix
• 7953e22 ci: upgrade GitHub Actions to use setup-node@v6 and checkout@v6
• 86dc1f8 Sanitizing input.
• See full diff in compare view

Updates form-data from 4.0.5 to 4.0.6

Changelog

Sourced from form-data's changelog.

v4.0.6 - 2026-06-12

Commits

• [Fix] escape CR, LF, and " in field names and filenames 8dff42c
• [Dev Deps] update @ljharb/eslint-config, auto-changelog, tape f31d21e
• [Deps] update hasown, mime-types 92ae0eb
• [Dev Deps] update js-randomness-predictor 67b0f65

Commits

• 64190db v4.0.6
• 92ae0eb [Deps] update hasown, mime-types
• f31d21e [Dev Deps] update @ljharb/eslint-config, auto-changelog, tape
• 8dff42c [Fix] escape CR, LF, and " in field names and filenames
• 67b0f65 [Dev Deps] update js-randomness-predictor
• See full diff in compare view

Updates undici from 6.26.0 to 6.27.0

Release notes

Sourced from undici's releases.

v8.5.0

⚠️ Security Release

This release line addresses 8 security advisories. Most are fixed in v8.5.0; the SOCKS5 pool-reuse issue was fixed earlier in v8.2.0.

Action required: Upgrade to undici 8.5.0 or later.

npm install undici@^8.5.0

Summary

Advisory	CVE	Severity (CVSS)	Fixed in	Fix commit
GHSA-vxpw-j846-p89q	CVE-2026-12151	High (7.5)	8.5.0	32dbf0b3
GHSA-38rv-x7px-6hhq	CVE-2026-9675	High (7.5)	8.5.0	b4c287b3
GHSA-vmh5-mc38-953g	CVE-2026-9697	High (7.4)	8.5.0	42d49559
GHSA-hm92-r4w5-c3mj	CVE-2026-6734	High (7.5)	8.2.0	a516f870
GHSA-pr7r-676h-xcf6	CVE-2026-9678	Moderate (5.9)	8.5.0	cb105d7c
GHSA-p88m-4jfj-68fv	CVE-2026-9679	Moderate (5.9)	8.5.0	5655ea43
GHSA-g8m3-5g58-fq7m	CVE-2026-11525	Low (3.7)	8.5.0	5655ea43
GHSA-35p6-xmwp-9g52	CVE-2026-6733	Low (3.7)	8.5.0	6ea54ef8

---

High severity

WebSocket DoS via fragment count bypass — CVE-2026-12151

GHSA-vxpw-j846-p89q · CWE-400, CWE-770 Fix: 32dbf0b3 websocket: limit the number of fragments in a message (also c5ed7875 handle empty fragments and stream limits)

A malicious WebSocket server can stream a large number of small or empty continuation frames. Undici enforced a limit on cumulative payload size but did not limit the number of fragments per message, leading to unbounded memory growth and denial of service.

• Affected: applications using new WebSocket(...) or WebSocketStream against untrusted endpoints.
• Workaround: none — upgrade is required.

WebSocket DoS via cumulative fragment bypass — CVE-2026-9675

GHSA-38rv-x7px-6hhq · CWE-400, CWE-770 Fix: b4c287b3 fix(websocket): enforce max payload size across fragments

Undici validated the size of individual frames but did not track cumulative size across a fragmented message. An attacker could send many small fragments that each pass per-frame validation but collectively exceed the configured limit, causing memory exhaustion. This is a regression introduced in 8.1.0 (the

... (truncated)

Commits

• a0806e1 Bumped v8.5.0 (#5429)
• 8a0392c test: detect available python command in wpt runner (#5427)
• f4045b9 ci: increase Node.js workflow timeout (#5426)
• 363e44f chore: removed repro-h2-pipelining-default.mjs and lint (#5420)
• c5ed787 websocket: handle empty fragments and stream limits
• e114e77 align EventSource with spec (#5418)
• 6df53c5 fix: preserve h2 queue on out-of-order completion (#5410)
• 32dbf0b websocket: limit the number of fragments in a message
• 0d6ecc5 add bodymixin.textStream() (#5416)
• 42d4955 fix: honor requestTls when proxy is SOCKS5
• Additional commits viewable in compare view

Updates undici from 8.3.0 to 8.5.0

Release notes

Sourced from undici's releases.

v8.5.0

⚠️ Security Release

This release line addresses 8 security advisories. Most are fixed in v8.5.0; the SOCKS5 pool-reuse issue was fixed earlier in v8.2.0.

Action required: Upgrade to undici 8.5.0 or later.

npm install undici@^8.5.0

Summary

Advisory	CVE	Severity (CVSS)	Fixed in	Fix commit
GHSA-vxpw-j846-p89q	CVE-2026-12151	High (7.5)	8.5.0	32dbf0b3
GHSA-38rv-x7px-6hhq	CVE-2026-9675	High (7.5)	8.5.0	b4c287b3
GHSA-vmh5-mc38-953g	CVE-2026-9697	High (7.4)	8.5.0	42d49559
GHSA-hm92-r4w5-c3mj	CVE-2026-6734	High (7.5)	8.2.0	a516f870
GHSA-pr7r-676h-xcf6	CVE-2026-9678	Moderate (5.9)	8.5.0	cb105d7c
GHSA-p88m-4jfj-68fv	CVE-2026-9679	Moderate (5.9)	8.5.0	5655ea43
GHSA-g8m3-5g58-fq7m	CVE-2026-11525	Low (3.7)	8.5.0	5655ea43
GHSA-35p6-xmwp-9g52	CVE-2026-6733	Low (3.7)	8.5.0	6ea54ef8

---

High severity

WebSocket DoS via fragment count bypass — CVE-2026-12151

GHSA-vxpw-j846-p89q · CWE-400, CWE-770 Fix: 32dbf0b3 websocket: limit the number of fragments in a message (also c5ed7875 handle empty fragments and stream limits)

A malicious WebSocket server can stream a large number of small or empty continuation frames. Undici enforced a limit on cumulative payload size but did not limit the number of fragments per message, leading to unbounded memory growth and denial of service.

• Affected: applications using new WebSocket(...) or WebSocketStream against untrusted endpoints.
• Workaround: none — upgrade is required.

WebSocket DoS via cumulative fragment bypass — CVE-2026-9675

GHSA-38rv-x7px-6hhq · CWE-400, CWE-770 Fix: b4c287b3 fix(websocket): enforce max payload size across fragments

Undici validated the size of individual frames but did not track cumulative size across a fragmented message. An attacker could send many small fragments that each pass per-frame validation but collectively exceed the configured limit, causing memory exhaustion. This is a regression introduced in 8.1.0 (the

... (truncated)

Commits

• a0806e1 Bumped v8.5.0 (#5429)
• 8a0392c test: detect available python command in wpt runner (#5427)
• f4045b9 ci: increase Node.js workflow timeout (#5426)
• 363e44f chore: removed repro-h2-pipelining-default.mjs and lint (#5420)
• c5ed787 websocket: handle empty fragments and stream limits
• e114e77 align EventSource with spec (#5418)
• 6df53c5 fix: preserve h2 queue on out-of-order completion (#5410)
• 32dbf0b websocket: limit the number of fragments in a message
• 0d6ecc5 add bodymixin.textStream() (#5416)
• 42d4955 fix: honor requestTls when proxy is SOCKS5
• Additional commits viewable in compare view

Updates esbuild from 0.28.0 to 0.28.1

Release notes

Sourced from esbuild's releases.

v0.25.12

• Fix a minification regression with CSS media queries (#4315)

  The previous release introduced support for parsing media queries which unintentionally introduced a regression with the removal of duplicate media rules during minification. Specifically the grammar for @media <media-type> and <media-condition-without-or> { ... } was missing an equality check for the <media-condition-without-or> part, so rules with different suffix clauses in this position would incorrectly compare equal and be deduplicated. This release fixes the regression.

• Update the list of known JavaScript globals (#4310)

  This release updates esbuild's internal list of known JavaScript globals. These are globals that are known to not have side-effects when the property is accessed. For example, accessing the global Array property is considered to be side-effect free but accessing the global scrollY property can trigger a layout, which is a side-effect. This is used by esbuild's tree-shaking to safely remove unused code that is known to be side-effect free. This update adds the following global properties:

  From ES2017:

  • Atomics
  • SharedArrayBuffer

  From ES2020:

  • BigInt64Array
  • BigUint64Array

  From ES2021:

  • FinalizationRegistry
  • WeakRef

  From ES2025:

  • Float16Array
  • Iterator

  Note that this does not indicate that constructing any of these objects is side-effect free, just that accessing the identifier is side-effect free. For example, this now allows esbuild to tree-shake classes that extend from Iterator:

  // This can now be tree-shaken by esbuild:
  class ExampleIterator extends Iterator {}

• Add support for the new @view-transition CSS rule (#4313)

  With this release, esbuild now has improved support for pretty-printing and minifying the new @view-transition rule (which esbuild was previously unaware of):

  /* Original code */
  @view-transition {
    navigation: auto;
    types: check;
  }
  /* Old output */
  @​view-transition { navigation: auto; types: check; }
  /* New output */
  @​view-transition {
  navigation: auto;
  types: check;

... (truncated)

Changelog

Sourced from esbuild's changelog.

0.25.12

• Fix a minification regression with CSS media queries (#4315)

  The previous release introduced support for parsing media queries which unintentionally introduced a regression with the removal of duplicate media rules during minification. Specifically the grammar for @media <media-type> and <media-condition-without-or> { ... } was missing an equality check for the <media-condition-without-or> part, so rules with different suffix clauses in this position would incorrectly compare equal and be deduplicated. This release fixes the regression.

• Update the list of known JavaScript globals (#4310)

  This release updates esbuild's internal list of known JavaScript globals. These are globals that are known to not have side-effects when the property is accessed. For example, accessing the global Array property is considered to be side-effect free but accessing the global scrollY property can trigger a layout, which is a side-effect. This is used by esbuild's tree-shaking to safely remove unused code that is known to be side-effect free. This update adds the following global properties:

  From ES2017:

  • Atomics
  • SharedArrayBuffer

  From ES2020:

  • BigInt64Array
  • BigUint64Array

  From ES2021:

  • FinalizationRegistry
  • WeakRef

  From ES2025:

  • Float16Array
  • Iterator

  Note that this does not indicate that constructing any of these objects is side-effect free, just that accessing the identifier is side-effect free. For example, this now allows esbuild to tree-shake classes that extend from Iterator:

  // This can now be tree-shaken by esbuild:
  class ExampleIterator extends Iterator {}

• Add support for the new @view-transition CSS rule (#4313)

  With this release, esbuild now has improved support for pretty-printing and minifying the new @view-transition rule (which esbuild was previously unaware of):

  /* Original code */
  @view-transition {
    navigation: auto;
    types: check;
  }
  /* Old output */
  @​view-transition { navigation: auto; types: check; }
  /* New output */
  @​view-transition {
  navigation: auto;

... (truncated)

Commits

• 208f539 publish 0.25.12 to npm
• 5f03afd update release notes
• 6b2ee78 minify: remove css rules containing empty :is()
• f361deb add some additional known static methods
• 07aa646 automatically mark "RegExp.escape()" calls as pure
• 9039c46 simplify some call expression checks
• 188944d add some additional known static methods
• d3c67f9 fix #4310: add Iterator and other known globals
• 4a51f0b fix: escape dev server breadcrumb hrefs properly (#4316)
• 26b29ed fix #4315: @media deduplication bug edge case
• Additional commits viewable in compare view

Updates form-data from 4.0.5 to 4.0.6

Changelog

Sourced from form-data's changelog.

v4.0.6 - 2026-06-12

Commits

• [Fix] escape CR, LF, and " in field names and filenames 8dff42c
• [Dev Deps] update @ljharb/eslint-config, auto-changelog, tape f31d21e
• [Deps] update hasown, mime-types 92ae0eb
• [Dev Deps] update js-randomness-predictor 67b0f65

Commits

• 64190db v4.0.6
• 92ae0eb [Deps] update hasown, mime-types
• f31d21e [Dev Deps] update @ljharb/eslint-config, auto-changelog, tape
• 8dff42c [Fix] escape CR, LF, and " in field names and filenames
• 67b0f65 [Dev Deps] update js-randomness-predictor
• See full diff in compare view

Updates js-cookie from 2.2.1 to 3.0.8

Release notes

Sourced from js-cookie's releases.

v3.0.8

• Restore ES5 compatibility, inadvertently broken in 3.0.7 - #959
• Lift Node version restriction, inadvertently restricted to >= 20 in 3.0.7 - #956

v3.0.7

• Prevent cookie attribute injection: CVE-2026-46625 (eb3c40e)
• Add Partitioned attribute to readme (b994768)
• Publish to npm registry via trusted publisher exclusively (4dc71be)
• Ensure consistent behaviour for get('name') + get() (1953d30)

v3.0.5

• Remove npm version restriction in package.json - #818

v3.0.4

• Publish to npmjs.com with package provenance

v3.0.1

• Make package.json accessible in export - #727

v3.0.0

• Removed defaults in favor of a builder: now to supply an api instance with particular predefined (cookie) attributes there's Cookies.withAttributes(), e.g.:

const api = Cookies.withAttributes({
  path: '/',
  secure: true
})
api.set('key', 'value') // writes cookie with path: '/' and secure: true...

• The attributes that an api instance is configured with are exposed as attributes property; it's an immutable object and unlike defaults cannot be changed to configure the api.
• The mechanism to fall back to the standard, internal converter by returning a falsy value in a custom read converter has been removed. Instead the default converters are now exposed as Cookies.converter, which allows for implementing self-contained custom converters providing the same behavior:

const customReadConverter = (value, name) => {
  if (name === 'special') {
    return unescape(value)
  }
  return Cookies.converter.read(value)
}

• withConverter() no longer accepts a function as argument to be turned into a read converter. It is now required to always pass an object with the explicit type(s) of converter(s):

const api = Cookies.withConverter({
  read: (value, name) => unescape(value)
})

• The converter(s) that an api instance is configured with are exposed as converter property; it's an immutable object and cannot be changed to configure the api.
• Started providing library as ES module, in addition to UMD module. The module field in package.json points to an ES module variant of the library.
• Started using browser field instead of main in package.json (for the UMD variant of the library).
• Dropped support for IE < 10.
• Removed built-in JSON support, i.e. getJSON() and automatic stringifying in set(): use Cookies.set('foo', JSON.stringify({ ... })) and JSON.parse(Cookies.get('foo')) instead.
• Removed support for Bower.

... (truncated)

Commits

• d7a1096 Craft v3.0.8 release
• 248e685 Use existing Chrome with puppeteer
• fc04269 Remove QUnit related workaround in Grunt config
• 265a685 Tidy up package lock file
• 478e591 Disable Node deprecation DEP0044 for release workflow
• 331d524 Fix node version config for E2E test job
• 11d773d Ensure ECMAScript compatibility
• d788646 Remove engines property from package
• e7d9a4d Fix typo in test assertion message
• b5fca24 Make credentials use explicit in release workflow
• Additional commits viewable in compare view

Maintainer changes

This version was pushed to npm by GitHub Actions, a new releaser for js-cookie since your current version.

Updates ws from 8.20.1 to 8.21.0

Release notes

Sourced from ws's releases.

8.21.0

Features

• Introduced the maxBufferedChunks and maxFragments options (2b2abd45).

Bug fixes

• Fixed a remote memory exhaustion DoS vulnerability (2b2abd45).

A high volume of tiny fragments and data chunks could be sent by a peer, using modest network traffic, to crash a ws server or client due to OOM.

import { WebSocket, WebSocketServer } from 'ws';
const wss = new WebSocketServer({ port: 0 }, function () {
const data = Buffer.alloc(1);
const options = { fin: false };
const { port } = wss.address();
const ws = new WebSocket(ws://localhost:${port});
ws.on('open', function () {
(function send() {
ws.send(data, options, function (err) {
if (err) return;
send();
});
})();
});
ws.on('error', console.error);
ws.on('close', function (code, reason) {
console.log(client close - code: ${code} reason: ${reason.toString()});
});
});
wss.on('connection', function (ws) {
ws.on('error', console.error);
ws.on('close', function (code, reason) {
console.log(server close - code: ${code} reason: ${reason.toString()});
});
});

The vulnerability was responsibly disclosed and fixed by Nadav Magier.

In vulnerable versions, the issue can be mitigated by lowering the value of the maxPayload option if possible.

Commits

• bca91ad [dist] 8.21.0
• 2b2abd4 [security] Limit retained message parts
• 78eabe2 [security] Add latest vulnerability to SECURITY.md
• See full diff in compare view

Updates undici from 6.26.0 to 6.27.0

Release notes

Sourced from undici's releases.

v8.5.0

⚠️ Security Release

This release line addresses 8 security advisories. Most are fixed in v8.5.0; the SOCKS5 pool-reuse issue was fixed earlier in v8.2.0.

Action required: Upgrade to undici 8.5.0 or later.

npm install undici@^8.5.0

Summary

Advisory	CVE	Severity (CVSS)	Fixed in	Fix commit
GHSA-vxpw-j846-p89q	CVE-2026-12151	High (7.5)	8.5.0	32dbf0b3
GHSA-38rv-x7px-6hhq	CVE-2026-9675	High (7.5)	8.5.0	b4c287b3
GHSA-vmh5-mc38-953g	CVE-2026-9697	High (7.4)	8.5.0	42d49559
GHSA-hm92-r4w5-c3mj	CVE-2026-6734	High (7.5)	8.2.0	a516f870
GHSA-pr7r-676h-xcf6	CVE-2026-9678	Moderate (5.9)	8.5.0	cb105d7c
GHSA-p88m-4jfj-68fv	CVE-2026-9679	Moderate (5.9)	8.5.0	5655ea43
GHSA-g8m3-5g58-fq7m	CVE-2026-11525	Low (3.7)	8.5.0	5655ea43
GHSA-35p6-xmwp-9g52	CVE-2026-6733	Low (3.7)	8.5.0	6ea54ef8

---

High severity

WebSocket DoS via fragment count bypass — CVE-2026-12151

GHSA-vxpw-j846-p89q · CWE-400, CWE-770 Fix: 32dbf0b3 websocket: limit the number of fragments in a message (also c5ed7875 handle empty fragments and stream limits)

A malicious WebSocket server can stream a large number of small or empty continuation frames. Undici enforced a limit on cumulative payload size but did not limit the number of fragments per message, leading to unbounded memory growth and denial of service.

• Affected: applications using new WebSocket(...) or WebSocketStream against untrusted endpoints.
• Workaround: none — upgrade is required.

WebSocket DoS via cumulative fragment bypass — CVE-2026-9675

GHSA-38rv-x7px-6hhq · CWE-400, CWE-770 Fix: b4c287b3 fix(websocket): enforce max payload size across fragments

Undici validated the size of individual frames but did not track cumulative size across a fragmented message. An attacker could send many small fragments that each pass per-frame validation but collectively exceed the configured limit, causing memory exhaustion. This is a regression introduced in 8.1.0 (the

... (truncated)

Commits

• a0806e1 Bumped v8.5.0 (#5429)
• 8a0392c test: detect available python command in wpt runner (#5427)
• f4045b9 ci: increase Node.js workflow timeout (#5426)
• 363e44f chore: removed repro-h2-pipelining-default.mjs and lint (#5420)
• c5ed787 websocket: handle empty fragments and stream limits
• e114e77 align EventSource with spec (#5418)
• 6df53c5 fix: preserve h2 queue on out-of-order completion (#5410)
• 32dbf0b websocket: limit the number of fragments in a message
• 0d6ecc5 add bodymixin.textStream() (#5416)
• 42d4955 fix: honor requestTls when proxy is SOCKS5
• Additional commits viewable in compare view

Updates undici from 8.3.0 to 8.5.0

Release notes

Sourced from undici's releases.

v8.5.0

⚠️ Security Release

This release line addresses 8 security advisories. Most are fixed in v8.5.0; the SOCKS5 pool-reuse issue was fixed earlier in v8.2.0.

Action required: Upgrade to undici 8.5.0 or later.

npm install undici@^8.5.0

Summary

Advisory	CVE	Severity (CVSS)	Fixed in	Fix commit
GHSA-vxpw-j846-p89q	CVE-2026-12151	High (7.5)	8.5.0	32dbf0b3
GHSA-38rv-x7px-6hhq	CVE-2026-9675	High (7.5)	8.5.0	b4c287b3
GHSA-vmh5-mc38-953g	CVE-2026-9697	High (7.4)	8.5.0	42d49559
GHSA-hm92-r4w5-c3mj	CVE-2026-6734	High (7.5)	8.2.0	a516f870
GHSA-pr7r-676h-xcf6	CVE-2026-9678	Moderate (5.9)	8.5.0	cb105d7c
GHSA-p88m-4jfj-68fv	CVE-2026-9679	Moderate (5.9)	8.5.0	5655ea43
GHSA-g8m3-5g58-fq7m	CVE-2026-11525	Low (3.7)	8.5.0	5655ea43
GHSA-35p6-xmw... Description has been truncated

[socket-security]

Review the following changes in direct dependencies. Learn more about Socket for GitHub.

Diff	Package	Supply Chain Security	Vulnerability	Quality	Maintenance	License
	ethers@​6.16.0 ⏵ 6.17.0	+5
	ava@​8.0.1

View full report

[socket-security]

Caution

Review the following alerts detected in dependencies.

According to your organization's Security Policy, you must resolve all "Block" alerts before proceeding. Learn more about Socket for GitHub.

Action	Severity	Alert  (click "▶" to expand/collapse)
Block		Potential code anomaly (AI signal): npm cbor2 is 65.0% likely to have a medium risk anomaly Notes: Overall, the codebase appears to be a robust, feature-rich CBOR serialization library with extensive tag support and defensive input handling. There is no concrete evidence of malicious behavior or data exfiltration within the provided fragment. The main operational risk is inherent to any data-serialization component: untrusted inputs can trigger potentially expensive or error-prone constructions (URL/RegExp/Date) if downstream usage is insecure. The code mitigates this with strict validations and explicit error handling. Confidence: 0.65 Severity: 0.52 From: packages/plugin-hardhat/examples/BoxTransparent/package-lock.json → npm/ava@8.0.1 → npm/cbor2@2.3.0 ℹ Read more on: This package | This alert | What is an AI-detected potential code anomaly? Next steps: Take a moment to review the security alert above. Review the linked package source code to understand the potential risk. Ensure the package is not malicious before proceeding. If you're unsure how to proceed, reach out to your security team or ask the Socket team for help at support@socket.dev. Suggestion: An AI system found a low-risk anomaly in this package. It may still be fine to use, but you should check that it is safe before proceeding. Mark the package as acceptable risk. To ignore this alert only in this pull request, reply with the comment @SocketSecurity ignore npm/cbor2@2.3.0. You can also ignore all packages with @SocketSecurity ignore-all. To ignore an alert for all future pull requests, use Socket's Dashboard to change the triage state of this alert.
Block		Potential code anomaly (AI signal): npm esbuild is 68.0% likely to have a medium risk anomaly Notes: The esbuild install.js script downloads and installs platform-specific binaries through multiple paths (including direct tarballs) and lacks cryptographic integrity verification, creating supply-chain and execution risks. The installer’s use of a direct download/install and an override/wrapper mechanism, while potentially legitimate, heightens risk due to limited visibility and potential misuse; strict integrity checks and tighter auditing/logging are recommended. Confidence: 0.68 Severity: 0.60 From: ? → npm/hardhat@3.6.0 → npm/esbuild@0.25.12 ℹ Read more on: This package | This alert | What is an AI-detected potential code anomaly? Next steps: Take a moment to review the security alert above. Review the linked package source code to understand the potential risk. Ensure the package is not malicious before proceeding. If you're unsure how to proceed, reach out to your security team or ask the Socket team for help at support@socket.dev. Suggestion: An AI system found a low-risk anomaly in this package. It may still be fine to use, but you should check that it is safe before proceeding. Mark the package as acceptable risk. To ignore this alert only in this pull request, reply with the comment @SocketSecurity ignore npm/esbuild@0.25.12. You can also ignore all packages with @SocketSecurity ignore-all. To ignore an alert for all future pull requests, use Socket's Dashboard to change the triage state of this alert.
Block		Potential code anomaly (AI signal): npm esbuild is 90.0% likely to have a medium risk anomaly Notes: The esbuild package uses a postinstall install.js script to download platform-specific binaries from registry sources and verify them via hashes. While hash verification reduces risk, the elevated postinstall action creates a potential code-execution surface if the script is tampered with. Audit install.js and its endpoints, ensure artifacts are strictly verified against known hashes, and test in controlled environments before deployment. Confidence: 0.90 Severity: 0.60 From: packages/plugin-hardhat/examples/BoxTransparent/package-lock.json → npm/tsx@4.22.0 → npm/tsx@4.22.3 → npm/esbuild@0.28.1 ℹ Read more on: This package | This alert | What is an AI-detected potential code anomaly? Next steps: Take a moment to review the security alert above. Review the linked package source code to understand the potential risk. Ensure the package is not malicious before proceeding. If you're unsure how to proceed, reach out to your security team or ask the Socket team for help at support@socket.dev. Suggestion: An AI system found a low-risk anomaly in this package. It may still be fine to use, but you should check that it is safe before proceeding. Mark the package as acceptable risk. To ignore this alert only in this pull request, reply with the comment @SocketSecurity ignore npm/esbuild@0.28.1. You can also ignore all packages with @SocketSecurity ignore-all. To ignore an alert for all future pull requests, use Socket's Dashboard to change the triage state of this alert.
Block		Potential code anomaly (AI signal): npm glob is 68.0% likely to have a medium risk anomaly Notes: The Glob utilities implement a conventional and well-structured filesystem glob-walking mechanism with robust control flow (abort signals, backpressure) and safe output semantics. There is no evidence of malicious behavior, backdoors, or data exfiltration within this fragment. Risks mainly relate to how downstream consumers may handle emitted paths, not to the library itself. Confidence: 0.68 Severity: 0.50 From: packages/plugin-hardhat/examples/BoxTransparent/package-lock.json → npm/ava@8.0.1 → npm/glob@13.0.6 ℹ Read more on: This package | This alert | What is an AI-detected potential code anomaly? Next steps: Take a moment to review the security alert above. Review the linked package source code to understand the potential risk. Ensure the package is not malicious before proceeding. If you're unsure how to proceed, reach out to your security team or ask the Socket team for help at support@socket.dev. Suggestion: An AI system found a low-risk anomaly in this package. It may still be fine to use, but you should check that it is safe before proceeding. Mark the package as acceptable risk. To ignore this alert only in this pull request, reply with the comment @SocketSecurity ignore npm/glob@13.0.6. You can also ignore all packages with @SocketSecurity ignore-all. To ignore an alert for all future pull requests, use Socket's Dashboard to change the triage state of this alert.
Block		Potential code anomaly (AI signal): npm undici is 75.0% likely to have a medium risk anomaly Notes: The analyzed code appears to implement a standard in-memory cache batch operation flow (put/delete) with careful handling of response bodies by buffering and storing bytes for caching. No signs of malware, data exfiltration, backdoors, or obfuscated behavior were found. The primary security considerations relate to memory usage from buffering potentially large response bodies and ensuring robust validation within batch operations to prevent cache state corruption. Overall risk is moderate, driven by in-memory data handling rather than external communication. Confidence: 0.75 Severity: 0.60 From: packages/plugin-hardhat/examples/BoxTransparent/package-lock.json → npm/@nomicfoundation/hardhat-ethers@4.0.11 → npm/@nomicfoundation/hardhat-ethers@4.0.12 → npm/hardhat@3.6.0 → npm/undici@6.27.0 ℹ Read more on: This package | This alert | What is an AI-detected potential code anomaly? Next steps: Take a moment to review the security alert above. Review the linked package source code to understand the potential risk. Ensure the package is not malicious before proceeding. If you're unsure how to proceed, reach out to your security team or ask the Socket team for help at support@socket.dev. Suggestion: An AI system found a low-risk anomaly in this package. It may still be fine to use, but you should check that it is safe before proceeding. Mark the package as acceptable risk. To ignore this alert only in this pull request, reply with the comment @SocketSecurity ignore npm/undici@6.27.0. You can also ignore all packages with @SocketSecurity ignore-all. To ignore an alert for all future pull requests, use Socket's Dashboard to change the triage state of this alert.
Block		Potential code anomaly (AI signal): npm undici is 68.0% likely to have a medium risk anomaly Notes: The analyzed code implements a conventional HTTP/WebSocket-like upgrade handler with proper input validation, abort signal integration, and asynchronous callback management. It does not exhibit malicious activity such as data exfiltration or backdoors. The deliberate onHeaders error path is consistent with protocol expectations to reject non-upgrade responses. Overall security risk remains low to moderate, contingent on integration context, but no indicators of malware or obfuscation are detected in this fragment. Confidence: 0.68 Severity: 0.50 From: packages/plugin-hardhat/examples/BoxTransparent/package-lock.json → npm/@nomicfoundation/hardhat-ethers@4.0.11 → npm/@nomicfoundation/hardhat-ethers@4.0.12 → npm/hardhat@3.6.0 → npm/undici@6.27.0 ℹ Read more on: This package | This alert | What is an AI-detected potential code anomaly? Next steps: Take a moment to review the security alert above. Review the linked package source code to understand the potential risk. Ensure the package is not malicious before proceeding. If you're unsure how to proceed, reach out to your security team or ask the Socket team for help at support@socket.dev. Suggestion: An AI system found a low-risk anomaly in this package. It may still be fine to use, but you should check that it is safe before proceeding. Mark the package as acceptable risk. To ignore this alert only in this pull request, reply with the comment @SocketSecurity ignore npm/undici@6.27.0. You can also ignore all packages with @SocketSecurity ignore-all. To ignore an alert for all future pull requests, use Socket's Dashboard to change the triage state of this alert.
Block		Potential code anomaly (AI signal): npm undici is 65.0% likely to have a medium risk anomaly Notes: The code is a focused error-handling helper for HTTP responses that safely parses small payloads to include in an error object. It includes protective measures (chunk limits, controlled parsing, microtask-based callbacks) but uses unusual, brittle content-type checks and suppresses stack traces for debugging concealment. There is no evidence of malicious activity, data exfiltration, or backdoors within this fragment. The main risk is potential silent data loss if payloads exceed the chunk limit or mismatched content-type handling leads to missing payloads, but this is a functional trade-off rather than malicious. Suggested improvements include robust content-type parsing, clearer error signaling when payload is truncated, and optional logging to aid debugging without exposing stack traces in production. Confidence: 0.65 Severity: 0.58 From: packages/plugin-hardhat/examples/BoxTransparent/package-lock.json → npm/@nomicfoundation/hardhat-ethers@4.0.11 → npm/@nomicfoundation/hardhat-ethers@4.0.12 → npm/hardhat@3.6.0 → npm/undici@6.27.0 ℹ Read more on: This package | This alert | What is an AI-detected potential code anomaly? Next steps: Take a moment to review the security alert above. Review the linked package source code to understand the potential risk. Ensure the package is not malicious before proceeding. If you're unsure how to proceed, reach out to your security team or ask the Socket team for help at support@socket.dev. Suggestion: An AI system found a low-risk anomaly in this package. It may still be fine to use, but you should check that it is safe before proceeding. Mark the package as acceptable risk. To ignore this alert only in this pull request, reply with the comment @SocketSecurity ignore npm/undici@6.27.0. You can also ignore all packages with @SocketSecurity ignore-all. To ignore an alert for all future pull requests, use Socket's Dashboard to change the triage state of this alert.
Block		Potential code anomaly (AI signal): npm undici is 77.0% likely to have a medium risk anomaly Notes: The script performs an in-place, lossy re-encoding of a local file from UTF-8 to Latin-1 and rewrites it without backups or validation. This is unsafe due to potential data loss and code corruption, and could be exploited to tamper with source files in a supply chain. It does not exhibit active malware behavior, but its destructive nature warrants removal or strict safeguards (backups, explicit intent, error handling). Confidence: 0.77 Severity: 0.65 From: packages/plugin-hardhat/examples/BoxTransparent/package-lock.json → npm/@nomicfoundation/hardhat-ethers@4.0.11 → npm/@nomicfoundation/hardhat-ethers@4.0.12 → npm/hardhat@3.6.0 → npm/undici@6.27.0 ℹ Read more on: This package | This alert | What is an AI-detected potential code anomaly? Next steps: Take a moment to review the security alert above. Review the linked package source code to understand the potential risk. Ensure the package is not malicious before proceeding. If you're unsure how to proceed, reach out to your security team or ask the Socket team for help at support@socket.dev. Suggestion: An AI system found a low-risk anomaly in this package. It may still be fine to use, but you should check that it is safe before proceeding. Mark the package as acceptable risk. To ignore this alert only in this pull request, reply with the comment @SocketSecurity ignore npm/undici@6.27.0. You can also ignore all packages with @SocketSecurity ignore-all. To ignore an alert for all future pull requests, use Socket's Dashboard to change the triage state of this alert.
Block		Potential code anomaly (AI signal): npm undici is 63.0% likely to have a medium risk anomaly Notes: The file package/lib/llhttp/llhttp-wasm.js functions as a wrapper around an embedded WASM payload responsible for HTTP parsing, with obfuscated/low-level operations and lazy decoding that defers real behavior to the embedded binary. The lack of integrity checks and the embedded executable raise risk, and the true malicious intent cannot be confirmed without extracting and inspecting the WASM payload and how downstream code instantiates it. Confidence: 0.63 Severity: 0.55 From: packages/plugin-hardhat/examples/BoxTransparent/package-lock.json → npm/@openzeppelin/hardhat-upgrades@4.0.2 → npm/undici@8.5.0 ℹ Read more on: This package | This alert | What is an AI-detected potential code anomaly? Next steps: Take a moment to review the security alert above. Review the linked package source code to understand the potential risk. Ensure the package is not malicious before proceeding. If you're unsure how to proceed, reach out to your security team or ask the Socket team for help at support@socket.dev. Suggestion: An AI system found a low-risk anomaly in this package. It may still be fine to use, but you should check that it is safe before proceeding. Mark the package as acceptable risk. To ignore this alert only in this pull request, reply with the comment @SocketSecurity ignore npm/undici@8.5.0. You can also ignore all packages with @SocketSecurity ignore-all. To ignore an alert for all future pull requests, use Socket's Dashboard to change the triage state of this alert.
Block		Potential code anomaly (AI signal): npm undici is 68.0% likely to have a medium risk anomaly Notes: The code performs an in-place re-encoding of a local file (undici-fetch.js) and overwrites it with latin1-encoded data. There is no evidence of exfiltration, backdoors, or network activity. However, the lack of validation, error handling, and the fact that it can corrupt or permanently alter a source file constitutes a nontrivial risk. In a supply-chain or extension context, such a script could be misused to tamper with code. It is not inherently malicious by itself but is risky and should be restricted or audited before typical usage in a build or runtime environment. Confidence: 0.68 Severity: 0.60 From: packages/plugin-hardhat/examples/BoxTransparent/package-lock.json → npm/@openzeppelin/hardhat-upgrades@4.0.2 → npm/undici@8.5.0 ℹ Read more on: This package | This alert | What is an AI-detected potential code anomaly? Next steps: Take a moment to review the security alert above. Review the linked package source code to understand the potential risk. Ensure the package is not malicious before proceeding. If you're unsure how to proceed, reach out to your security team or ask the Socket team for help at support@socket.dev. Suggestion: An AI system found a low-risk anomaly in this package. It may still be fine to use, but you should check that it is safe before proceeding. Mark the package as acceptable risk. To ignore this alert only in this pull request, reply with the comment @SocketSecurity ignore npm/undici@8.5.0. You can also ignore all packages with @SocketSecurity ignore-all. To ignore an alert for all future pull requests, use Socket's Dashboard to change the triage state of this alert.

View full report

[dependabot]

Superseded by #1286.