GHSA-W2J7-F3C6-G8CW
Vulnerability from github – Published: 2026-06-23 22:46 – Updated: 2026-06-23 22:46Open Redirect in Flask-Security
Summary
flask_security.utils.validate_redirect_url() can allow an attacker-controlled redirect URL when subdomain redirects are enabled.
The bypass uses a backslash inside the URL authority/host:
http://evil.com\.whitelist.com
http://evil.com%5C.whitelist.com
Python's urlsplit() parses the full authority as evil.com\.whitelist.com or evil.com%5C.whitelist.com. Because the value ends with .whitelist.com, validate_redirect_url() accepts it as an allowed subdomain of whitelist.com.
This is similar in class to the previous Flask-Security-Too open redirect advisory CVE-2023-49438 / GHSA-672h-6x89-76m5, where crafted redirect URLs bypassed validation through browser URL normalization behavior.
Affected Configuration
The issue requires subdomain redirects to be enabled:
SERVER_NAME = "whitelist.com"
SECURITY_REDIRECT_ALLOW_SUBDOMAINS = True
Tested environment:
Flask-Security: 5.8.0
Flask: 3.1.3
Werkzeug: 3.1.8
Impact
An attacker can craft a URL that passes Flask-Security's redirect validation and produces a 302 response to an attacker-controlled URL-like authority.
This can be used for phishing or other attacks that rely on a trusted application redirecting users to an attacker-controlled destination.
Proof of Concept
PoC Flask App
from __future__ import annotations
from importlib.metadata import version
from urllib.parse import urlsplit
from flask import Flask, jsonify, redirect, request
from flask_security.utils import validate_redirect_url
app = Flask(__name__)
app.config.update(
SECRET_KEY="poc-only",
SERVER_NAME="whitelist.com",
SECURITY_REDIRECT_ALLOW_SUBDOMAINS=True,
SECURITY_REDIRECT_BASE_DOMAIN=None,
SECURITY_REDIRECT_ALLOWED_SUBDOMAINS=[],
)
@app.get("/")
def index():
return jsonify(
flask_version=version("Flask"),
configured_server_name=app.config["SERVER_NAME"],
examples=[
r"http://evil.com\.whitelist.com",
"http://evil.com%5C.whitelist.com",
"http://sub.whitelist.com",
"http://sub.not-whitelist.com",
],
)
@app.get("/check")
def check():
next_url = request.args.get("next", "")
parsed = urlsplit(next_url)
return jsonify(
next=next_url,
valid=validate_redirect_url(next_url),
parsed={
"scheme": parsed.scheme,
"netloc": parsed.netloc,
"hostname": parsed.hostname,
"path": parsed.path,
},
)
@app.get("/redir")
def redir():
next_url = request.args.get("next", "")
if not validate_redirect_url(next_url):
return jsonify(error="blocked", next=next_url), 400
return redirect(next_url)
if __name__ == "__main__":
app.run(host="127.0.0.1", port=5000, debug=False)
Steps to Reproduce
Run the PoC with the target project's Flask version:
.venv/bin/python poc_redirect_app.py
The invalid comparison case is correctly blocked:
http://127.0.0.1:5000/redir?next=http://evil.com
Observed result:
Check the validation result:
http://127.0.0.1:5000/check?next=http://evil.com%5C.whitelist.com
Observed result:
References
- CVE-2023-49438: https://advisories.gitlab.com/pypi/flask-security-too/CVE-2023-49438/
- GHSA-672h-6x89-76m5: https://osv.dev/vulnerability/CVE-2023-49438
- NVD entry for CVE-2023-49438: https://nvd.nist.gov/vuln/detail/CVE-2023-49438
{
"affected": [
{
"database_specific": {
"last_known_affected_version_range": "\u003c= 5.8.0"
},
"package": {
"ecosystem": "PyPI",
"name": "Flask-Security"
},
"ranges": [
{
"events": [
{
"introduced": "0"
},
{
"fixed": "5.8.1"
}
],
"type": "ECOSYSTEM"
}
]
}
],
"aliases": [],
"database_specific": {
"cwe_ids": [
"CWE-601"
],
"github_reviewed": true,
"github_reviewed_at": "2026-06-23T22:46:31Z",
"nvd_published_at": null,
"severity": "MODERATE"
},
"details": "# Open Redirect in Flask-Security\n\n## Summary\n\n`flask_security.utils.validate_redirect_url()` can allow an attacker-controlled redirect URL when subdomain redirects are enabled.\n\nThe bypass uses a backslash inside the URL authority/host:\n\n```text\nhttp://evil.com\\.whitelist.com\nhttp://evil.com%5C.whitelist.com\n```\n\nPython\u0027s `urlsplit()` parses the full authority as `evil.com\\.whitelist.com` or `evil.com%5C.whitelist.com`. Because the value ends with `.whitelist.com`, `validate_redirect_url()` accepts it as an allowed subdomain of `whitelist.com`.\n\nThis is similar in class to the previous Flask-Security-Too open redirect advisory CVE-2023-49438 / GHSA-672h-6x89-76m5, where crafted redirect URLs bypassed validation through browser URL normalization behavior.\n\n## Affected Configuration\n\nThe issue requires subdomain redirects to be enabled:\n\n```python\nSERVER_NAME = \"whitelist.com\"\nSECURITY_REDIRECT_ALLOW_SUBDOMAINS = True\n```\n\nTested environment:\n\n```text\nFlask-Security: 5.8.0\nFlask: 3.1.3\nWerkzeug: 3.1.8\n```\n\n## Impact\n\nAn attacker can craft a URL that passes Flask-Security\u0027s redirect validation and produces a `302` response to an attacker-controlled URL-like authority.\n\nThis can be used for phishing or other attacks that rely on a trusted application redirecting users to an attacker-controlled destination.\n\n## Proof of Concept\n\n### PoC Flask App\n\n```python\nfrom __future__ import annotations\n\nfrom importlib.metadata import version\nfrom urllib.parse import urlsplit\n\nfrom flask import Flask, jsonify, redirect, request\n\nfrom flask_security.utils import validate_redirect_url\n\n\napp = Flask(__name__)\napp.config.update(\n SECRET_KEY=\"poc-only\",\n SERVER_NAME=\"whitelist.com\",\n SECURITY_REDIRECT_ALLOW_SUBDOMAINS=True,\n SECURITY_REDIRECT_BASE_DOMAIN=None,\n SECURITY_REDIRECT_ALLOWED_SUBDOMAINS=[],\n)\n\n\n@app.get(\"/\")\ndef index():\n return jsonify(\n flask_version=version(\"Flask\"),\n configured_server_name=app.config[\"SERVER_NAME\"],\n examples=[\n r\"http://evil.com\\.whitelist.com\",\n \"http://evil.com%5C.whitelist.com\",\n \"http://sub.whitelist.com\",\n \"http://sub.not-whitelist.com\",\n ],\n )\n\n\n@app.get(\"/check\")\ndef check():\n next_url = request.args.get(\"next\", \"\")\n parsed = urlsplit(next_url)\n\n return jsonify(\n next=next_url,\n valid=validate_redirect_url(next_url),\n parsed={\n \"scheme\": parsed.scheme,\n \"netloc\": parsed.netloc,\n \"hostname\": parsed.hostname,\n \"path\": parsed.path,\n },\n )\n\n\n@app.get(\"/redir\")\ndef redir():\n next_url = request.args.get(\"next\", \"\")\n if not validate_redirect_url(next_url):\n return jsonify(error=\"blocked\", next=next_url), 400\n\n return redirect(next_url)\n\n\nif __name__ == \"__main__\":\n app.run(host=\"127.0.0.1\", port=5000, debug=False)\n```\n\n### Steps to Reproduce\n\nRun the PoC with the target project\u0027s Flask version:\n\n```bash\n.venv/bin/python poc_redirect_app.py\n```\n\nThe invalid comparison case is correctly blocked:\n\n```bash\nhttp://127.0.0.1:5000/redir?next=http://evil.com\n```\n\nObserved result:\n\n\u003cimg width=\"1019\" height=\"294\" alt=\"image\" src=\"https://github.com/user-attachments/assets/de25ac4d-b37f-4369-928e-f44dfd5b7557\" /\u003e\n\nCheck the validation result:\n\n```bash\nhttp://127.0.0.1:5000/check?next=http://evil.com%5C.whitelist.com\n```\n\nObserved result:\n\n\u003cimg width=\"1029\" height=\"634\" alt=\"image\" src=\"https://github.com/user-attachments/assets/8e5ec8a6-42a4-438a-8d12-a27724519091\" /\u003e\n\n## References\n\n- CVE-2023-49438: https://advisories.gitlab.com/pypi/flask-security-too/CVE-2023-49438/\n- GHSA-672h-6x89-76m5: https://osv.dev/vulnerability/CVE-2023-49438\n- NVD entry for CVE-2023-49438: https://nvd.nist.gov/vuln/detail/CVE-2023-49438",
"id": "GHSA-w2j7-f3c6-g8cw",
"modified": "2026-06-23T22:46:31Z",
"published": "2026-06-23T22:46:31Z",
"references": [
{
"type": "WEB",
"url": "https://github.com/pallets-eco/flask-security/security/advisories/GHSA-w2j7-f3c6-g8cw"
},
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2023-49438"
},
{
"type": "WEB",
"url": "https://advisories.gitlab.com/pypi/flask-security-too/CVE-2023-49438"
},
{
"type": "PACKAGE",
"url": "https://github.com/pallets-eco/flask-security"
},
{
"type": "WEB",
"url": "https://osv.dev/vulnerability/CVE-2023-49438"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:L/I:N/A:N",
"type": "CVSS_V3"
}
],
"summary": "Flask-Security has an Open Redirect issue"
}
Drag nodes to arrange them. Double-click a related vulnerability to open it, or a source to visit its link. The Physics button on the left tunes the layout and the simulation; the legend at the bottom left hides or shows nodes by kind and by sighting type.
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.
Vulnerability timeline
Sightings and recorded dates, grouped by day · UTC
How durations are calculated
Following the VTEM framework (Black Hat USA 2015), time to patch starts when the vendor or a coordinator was informed, total time to patch at discovery, and disclosure to patch at public disclosure; vendor response time runs from the report to its acknowledgement. Time to exploit and time unpatched start at the earliest discovery or report, falling back to public disclosure. Unpatched time ends at the first recorded patch, or the calculation time when no patch is recorded. Exploitation to patch shows the interval between those two events; a published proof of concept is a public exploit, not exploitation.
Each result shows its source dates. A negative interval means the end event came first; when date-only inputs leave the order open, the result says so. Date-only inputs are approximate. The earliest evidence wins: KEV report dates and sightings are labelled as such. Records describing another vulnerability are listed but never calculated from, and a PoC alone does not establish exploitation. Missing patch evidence does not prove that no patch exists. Filters below change the event list only.
Loading timeline…
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
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.