GCVE Workshop - 22 September 2026 (14:00-18:00), Luxembourg Before The Vulnopticon Conference - Registration

PYSEC-2026-353

Vulnerability from pysec - Published: 2026-06-29 11:50 - Updated: 2026-07-01 20:22
VLAI
Details

A vulnerability in h2oai/h2o-3 version 3.46.0.1 allows remote attackers to write arbitrary data to any file on the server. This is achieved by exploiting the /3/Parse endpoint to inject attacker-controlled data as the header of an empty file, which is then exported using the /3/Frames/framename/export endpoint. The impact of this vulnerability includes the potential for remote code execution and complete access to the system running h2o-3, as attackers can overwrite critical files such as private SSH keys or script files.

Severity
9.1 (Critical)
CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:H/A:H
Impacted products
Name purl
h2o pkg:pypi/h2o
Aliases
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "h2o",
        "purl": "pkg:pypi/h2o"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "last_affected": "3.46.0.1"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ],
      "versions": [
        "3.10.0.10",
        "3.10.0.3",
        "3.10.0.6",
        "3.10.0.7",
        "3.10.0.8",
        "3.10.3.3",
        "3.10.3.4",
        "3.10.4.1",
        "3.10.4.2",
        "3.10.4.3",
        "3.10.4.4",
        "3.10.4.6",
        "3.10.4.8",
        "3.16.0.1",
        "3.16.0.2",
        "3.16.0.3",
        "3.16.0.4",
        "3.18.0.1",
        "3.18.0.10",
        "3.18.0.11",
        "3.18.0.2",
        "3.18.0.3",
        "3.18.0.4",
        "3.18.0.5",
        "3.18.0.6",
        "3.18.0.7",
        "3.18.0.8",
        "3.18.0.9",
        "3.20.0.4",
        "3.20.0.5",
        "3.20.0.6",
        "3.20.0.7",
        "3.20.0.8",
        "3.22.0.1",
        "3.22.0.2",
        "3.22.0.3",
        "3.22.0.4",
        "3.22.0.5",
        "3.22.1.1",
        "3.22.1.2",
        "3.22.1.3",
        "3.22.1.4",
        "3.22.1.5",
        "3.22.1.6",
        "3.24.0.1",
        "3.24.0.2",
        "3.24.0.3",
        "3.24.0.4",
        "3.24.0.5",
        "3.26.0.1",
        "3.26.0.10",
        "3.26.0.11",
        "3.26.0.2",
        "3.26.0.3",
        "3.26.0.4",
        "3.26.0.5",
        "3.26.0.6",
        "3.26.0.8",
        "3.26.0.9",
        "3.28.0.1",
        "3.28.0.2",
        "3.28.0.3",
        "3.28.1.2",
        "3.28.1.3",
        "3.30.0.1",
        "3.30.0.2",
        "3.30.0.3",
        "3.30.0.4",
        "3.30.0.5",
        "3.30.0.6",
        "3.30.0.7",
        "3.30.1.1",
        "3.30.1.2",
        "3.30.1.3",
        "3.32.0.2",
        "3.32.0.3",
        "3.32.0.4",
        "3.32.0.5",
        "3.32.1.1",
        "3.32.1.2",
        "3.32.1.3",
        "3.32.1.4",
        "3.32.1.5",
        "3.32.1.6",
        "3.32.1.7",
        "3.34.0.3",
        "3.34.0.7",
        "3.34.0.8",
        "3.36.0.2",
        "3.36.0.3",
        "3.36.0.4",
        "3.36.1.1",
        "3.36.1.2",
        "3.36.1.3",
        "3.36.1.4",
        "3.36.1.5",
        "3.38.0.1",
        "3.38.0.2",
        "3.38.0.3",
        "3.38.0.4",
        "3.40.0.1",
        "3.40.0.2",
        "3.40.0.3",
        "3.40.0.4",
        "3.42.0.1",
        "3.42.0.2",
        "3.42.0.3",
        "3.42.0.4",
        "3.44.0.1",
        "3.44.0.2",
        "3.44.0.3",
        "3.46.0.1"
      ]
    }
  ],
  "aliases": [
    "CVE-2024-5986",
    "GHSA-wj3h-wx8g-x699"
  ],
  "details": "A vulnerability in h2oai/h2o-3 version 3.46.0.1 allows remote attackers to write arbitrary data to any file on the server. This is achieved by exploiting the `/3/Parse` endpoint to inject attacker-controlled data as the header of an empty file, which is then exported using the `/3/Frames/framename/export` endpoint. The impact of this vulnerability includes the potential for remote code execution and complete access to the system running h2o-3, as attackers can overwrite critical files such as private SSH keys or script files.",
  "id": "PYSEC-2026-353",
  "modified": "2026-07-01T20:22:54.140956Z",
  "published": "2026-06-29T11:50:51.728938Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2024-5986"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/h2oai/h2o-3"
    },
    {
      "type": "WEB",
      "url": "https://huntr.com/bounties/64ff5319-6ac3-4447-87f7-b53495d4d5a3"
    },
    {
      "type": "PACKAGE",
      "url": "https://pypi.org/project/h2o"
    },
    {
      "type": "ADVISORY",
      "url": "https://github.com/advisories/GHSA-wj3h-wx8g-x699"
    }
  ],
  "severity": [
    {
      "score": "CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:H/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "H2O has an External Control of File Name or Path vulnerability"
}



Log in or create an account to share your comment.




Tags
Taxonomy of the tags.





Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.

Sightings

Author Source Type Date Other

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.


Detection rules are retrieved from Rulezet.


Related by attack behaviour

Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.