GHSA-GCWF-C25F-P9RV

Vulnerability from github – Published: 2026-11-09 23:24 – Updated: 2026-11-09 23:24
VLAI
Summary
Kiwi TCMS account confirmation open redirect through protocol-relative URLs
Details

Summary

In Kiwi TCMS version 16.1, the fix for CVE-2026-54724 (open redirect via unvalidated next parameter) in the account confirmation endpoint is incomplete. The validation uses startswith("/") to check if the redirect URL is local, but this is bypassed by protocol-relative URLs like //evil.com which start with / but redirect the browser to an external attacker-controlled domain. 16.1’s incomplete mitigation still allows protocol-relative destinations and earlier releases lack the mitigation.

Vulnerable Code

In tcms/kiwi_auth/views.py, the Confirm view (line ~113):

def get_redirect_url(self, *args, **kwargs):
    self.url = self.request.GET.get("next", reverse("core-views-index"))
    if not self.url.startswith("/"):
        self.url = reverse("core-views-index")

The check startswith("/") passes for //evil.com because it begins with /, but browsers interpret //evil.com as a protocol-relative URL that redirects to https://evil.com (or http://evil.com depending on the current page's protocol).

Proof of Concept

GET /confirm/<activation_key>/?next=//evil.com/phishing HTTP/1.1
Host: kiwi.example.com

Response:

HTTP/1.1 302 Found
Location: //evil.com/phishing

The browser follows the redirect to https://evil.com/phishing.

Verification:

>>> url = "//evil.com/phishing"
>>> url.startswith("/")
True  # passes the check!

Impact

An attacker can craft a confirmation link that, after the user confirms their account, redirects them to a phishing page. This is useful for:

  • Credential phishing (fake login page on attacker domain)
  • Malware delivery after trusted action
  • Social engineering (user trusts the link because it's from the legitimate Kiwi TCMS instance)

The attack requires sending a crafted activation link to a user who is registering, or intercepting/predicting the activation key.

Suggested Fix

Use Django's url_has_allowed_host_and_scheme() utility instead of a simple startswith check:

from django.utils.http import url_has_allowed_host_and_scheme

def get_redirect_url(self, *args, **kwargs):
    self.url = self.request.GET.get("next", reverse("core-views-index"))
    if not url_has_allowed_host_and_scheme(
        self.url,
        allowed_hosts={self.request.get_host()},
        require_https=self.request.is_secure(),
    ):
        self.url = reverse("core-views-index")

This properly rejects protocol-relative URLs, absolute URLs to other hosts, and other bypass patterns.

Fixed in Version 16.2

Show details on source website

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "kiwitcms"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "last_affected": "12.4"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-77433"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-601"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-11-09T23:24:14Z",
    "nvd_published_at": null,
    "severity": "MODERATE"
  },
  "details": "## Summary\n\nIn Kiwi TCMS version 16.1, the fix for CVE-2026-54724 (open redirect via unvalidated `next` parameter) in the account confirmation endpoint is incomplete. The validation uses `startswith(\"/\")` to check if the redirect URL is local, but this is bypassed by protocol-relative URLs like `//evil.com` which start with `/` but redirect the browser to an external attacker-controlled domain. 16.1\u2019s incomplete mitigation still allows protocol-relative destinations and earlier releases lack the mitigation.\n\n## Vulnerable Code\n\nIn `tcms/kiwi_auth/views.py`, the `Confirm` view (line ~113):\n\n    def get_redirect_url(self, *args, **kwargs):\n        self.url = self.request.GET.get(\"next\", reverse(\"core-views-index\"))\n        if not self.url.startswith(\"/\"):\n            self.url = reverse(\"core-views-index\")\n\nThe check `startswith(\"/\")` passes for `//evil.com` because it begins with `/`, but browsers interpret `//evil.com` as a protocol-relative URL that redirects to `https://evil.com` (or `http://evil.com` depending on the current page\u0027s protocol).\n\n## Proof of Concept\n\n    GET /confirm/\u003cactivation_key\u003e/?next=//evil.com/phishing HTTP/1.1\n    Host: kiwi.example.com\n\nResponse:\n\n    HTTP/1.1 302 Found\n    Location: //evil.com/phishing\n\nThe browser follows the redirect to `https://evil.com/phishing`.\n\nVerification:\n\n    \u003e\u003e\u003e url = \"//evil.com/phishing\"\n    \u003e\u003e\u003e url.startswith(\"/\")\n    True  # passes the check!\n\n## Impact\n\nAn attacker can craft a confirmation link that, after the user confirms their account, redirects them to a phishing page. This is useful for:\n\n- Credential phishing (fake login page on attacker domain)\n- Malware delivery after trusted action\n- Social engineering (user trusts the link because it\u0027s from the legitimate Kiwi TCMS instance)\n\nThe attack requires sending a crafted activation link to a user who is registering, or intercepting/predicting the activation key.\n\n## Suggested Fix\n\nUse Django\u0027s `url_has_allowed_host_and_scheme()` utility instead of a simple `startswith` check:\n\n    from django.utils.http import url_has_allowed_host_and_scheme\n\n    def get_redirect_url(self, *args, **kwargs):\n        self.url = self.request.GET.get(\"next\", reverse(\"core-views-index\"))\n        if not url_has_allowed_host_and_scheme(\n            self.url,\n            allowed_hosts={self.request.get_host()},\n            require_https=self.request.is_secure(),\n        ):\n            self.url = reverse(\"core-views-index\")\n\nThis properly rejects protocol-relative URLs, absolute URLs to other hosts, and other bypass patterns.\n\nFixed in Version 16.2",
  "id": "GHSA-gcwf-c25f-p9rv",
  "modified": "2026-11-09T23:24:14Z",
  "published": "2026-11-09T23:24:14Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/kiwitcms/Kiwi/security/advisories/GHSA-gcwf-c25f-p9rv"
    },
    {
      "type": "WEB",
      "url": "https://github.com/kiwitcms/Kiwi/security/advisories/GHSA-hmj5-jm8h-h9fh"
    },
    {
      "type": "WEB",
      "url": "https://github.com/kiwitcms/Kiwi/pull/4458"
    },
    {
      "type": "WEB",
      "url": "https://github.com/kiwitcms/Kiwi/commit/65d8f7c2743d7f06a4fca77c060b9d561f02f948"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/kiwitcms/Kiwi"
    },
    {
      "type": "WEB",
      "url": "https://github.com/kiwitcms/Kiwi/releases/tag/v16.2"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Kiwi TCMS account confirmation open redirect through protocol-relative URLs"
}



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…

Loading…

Loading…

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.


Loading…