PYSEC-2026-174

Vulnerability from pysec - Published: 2026-06-01 09:16 - Updated: 2026-06-02 10:33
VLAI
Details

Exploitation requires the attacker to already be an authenticated Airflow worker holding a valid Log-server JWT issued for at least one Dag. Apache Airflow's Log server authorized JWT tokens against Dag IDs by applying Python's str.lstrip() to the requested path segment when verifying the JWT's sub claim. str.lstrip() strips any of a set of characters from the left (not a prefix), so a JWT issued for a Dag named e.g. dag_a would authorize log access to any other Dag whose name began with any subset of the characters {d, a, g, _} (e.g. dag_attacker, aaaa_target, _dag_secret). Such an authenticated worker could enumerate and read worker logs of other Dags whose names happened to share that character-class prefix, leaking task output and error traces beyond the documented per-Dag isolation boundary. Affects deployments relying on per-Dag log-access scoping (multi-team, shared-executor, shared-worker topologies). Users are advised to upgrade to apache-airflow 3.2.2 or later.

Impacted products
Name purl
apache-airflow pkg:pypi/apache-airflow
Aliases

{
  "affected": [
    {
      "ecosystem_specific": {},
      "package": {
        "ecosystem": "PyPI",
        "name": "apache-airflow",
        "purl": "pkg:pypi/apache-airflow"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "3.0.0"
            },
            {
              "fixed": "3.2.2"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ],
      "versions": [
        "3.0.0",
        "3.0.1",
        "3.0.1rc1",
        "3.0.2",
        "3.0.2rc1",
        "3.0.2rc2",
        "3.0.3",
        "3.0.3rc1",
        "3.0.3rc2",
        "3.0.3rc3",
        "3.0.3rc4",
        "3.0.3rc5",
        "3.0.3rc6",
        "3.0.4",
        "3.0.4rc1",
        "3.0.4rc2",
        "3.0.5",
        "3.0.5rc1",
        "3.0.5rc2",
        "3.0.5rc3",
        "3.0.6",
        "3.0.6rc1",
        "3.0.6rc2",
        "3.1.0",
        "3.1.0b1",
        "3.1.0b2",
        "3.1.0rc1",
        "3.1.0rc2",
        "3.1.1",
        "3.1.1rc1",
        "3.1.1rc2",
        "3.1.2",
        "3.1.2rc1",
        "3.1.2rc2",
        "3.1.3",
        "3.1.3rc1",
        "3.1.4",
        "3.1.4rc1",
        "3.1.4rc2",
        "3.1.5",
        "3.1.5rc1",
        "3.1.6",
        "3.1.6rc1",
        "3.1.7",
        "3.1.7rc1",
        "3.1.7rc2",
        "3.1.8",
        "3.1.8rc1",
        "3.1.8rc2",
        "3.2.0",
        "3.2.0b1",
        "3.2.0b2",
        "3.2.0rc1",
        "3.2.0rc2",
        "3.2.1",
        "3.2.1rc1",
        "3.2.1rc2",
        "3.2.1rc3",
        "3.2.2rc1",
        "3.2.2rc2",
        "3.2.2rc3"
      ]
    }
  ],
  "aliases": [
    "CVE-2026-45426"
  ],
  "details": "Exploitation requires the attacker to already be an authenticated Airflow worker holding a valid Log-server JWT issued for at least one Dag. Apache Airflow\u0027s Log server authorized JWT tokens against Dag IDs by applying Python\u0027s `str.lstrip()` to the requested path segment when verifying the JWT\u0027s `sub` claim. `str.lstrip()` strips any of a *set* of characters from the left (not a prefix), so a JWT issued for a Dag named e.g. `dag_a` would authorize log access to any other Dag whose name began with any subset of the characters `{d, a, g, _}` (e.g. `dag_attacker`, `aaaa_target`, `_dag_secret`). Such an authenticated worker could enumerate and read worker logs of other Dags whose names happened to share that character-class prefix, leaking task output and error traces beyond the documented per-Dag isolation boundary. Affects deployments relying on per-Dag log-access scoping (multi-team, shared-executor, shared-worker topologies). Users are advised to upgrade to `apache-airflow` 3.2.2 or later.",
  "id": "PYSEC-2026-174",
  "modified": "2026-06-02T10:33:39.263217Z",
  "published": "2026-06-01T09:16:19.583Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "http://www.openwall.com/lists/oss-security/2026/05/31/13"
    },
    {
      "type": "ADVISORY",
      "url": "https://lists.apache.org/thread/hz1q7vg65vq2h4fobv5ww8tp257fbqj9"
    },
    {
      "type": "FIX",
      "url": "https://github.com/apache/airflow/pull/66749"
    }
  ],
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:L/I:N/A:N",
      "type": "CVSS_V3"
    }
  ]
}


Log in or create an account to share your comment.




Tags
Taxonomy of the tags.


Loading…

Loading…

Loading…

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.

Loading…

Detection rules are retrieved from Rulezet.

Loading…

Loading…