Summary
Unfurl's debug mode cannot be disabled due to string config parsing (Werkzeug debugger exposure)
The Unfurl web app enables Flask debug mode even when configuration sets debug = False. The config value is read as a string and passed directly to app.run(debug=...), so any non-empty string evaluates truthy. This leaves the Werkzeug debugger active by default.
Details
unfurl/app.py:web_app()readsdebugviaconfig['UNFURL_APP'].get('debug'), which returns a string.UnfurlApp.__init__passes that string directly toapp.run(debug=unfurl_debug, ...).- If
unfurl.iniomitsdebug, the default argument is the string"True". - As a result, debug mode is effectively always on and cannot be reliably disabled via config.
PoC
- Create a local
unfurl.iniwithdebug = Falseunder[UNFURL_APP]. - Run the server using
unfurl_app(orpython -c 'from unfurl.app import web_app; web_app()'). - Observe server logs showing
Debug mode: on/Debugger is active!. - The included PoC script
security_poc/poc_debug_mode.py --spawnautomates this check.
PoC Script (inline)
#!/usr/bin/env python3
"""
Unfurl Debug Mode PoC (Corrected)
================================
This PoC demonstrates that Unfurl's Flask debug mode is effectively
**always enabled by default** due to string parsing of the `debug`
config value. Even `debug = False` in `unfurl.ini` evaluates truthy
when passed to `app.run(debug=...)`.
Two modes:
1) --spawn (default): launch a local Unfurl server with debug=False
in a temp config and inspect logs for "Debug mode: on".
2) --target: attempt a remote indicator check (best-effort; may be silent
if no exception is triggered).
"""
import argparse
import os
import subprocess
import sys
import tempfile
import textwrap
import time
def run_spawn_check() -> None:
repo_root = os.path.abspath(os.path.join(os.path.dirname(__file__), '..'))
ini_contents = textwrap.dedent("""
[UNFURL_APP]
host = 127.0.0.1
port = 5055
debug = False
remote_lookups = false
[API_KEYS]
bitly =
macaddress_io =
""").strip() + "\n"
with tempfile.TemporaryDirectory() as tmp:
ini_path = os.path.join(tmp, 'unfurl.ini')
with open(ini_path, 'w') as f:
f.write(ini_contents)
env = os.environ.copy()
env['PYTHONPATH'] = repo_root
cmd = [sys.executable, '-c', 'from unfurl.app import web_app; web_app()']
proc = subprocess.Popen(
cmd,
cwd=tmp,
env=env,
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
text=True
)
# Allow server to start and emit logs
time.sleep(2)
proc.terminate()
try:
out, err = proc.communicate(timeout=2)
except subprocess.TimeoutExpired:
proc.kill()
out, err = proc.communicate()
output = (out or "") + (err or "")
print("\n[+] Debug mode spawn check")
print(" Config: debug = False")
if "Debug mode: on" in output or "Debugger is active" in output:
print(" ✅ Debug mode is ON despite debug=False (vulnerable)")
else:
print(" ⚠️ Debug mode not detected in logs (check output below)")
if output.strip():
print("\n--- server output (truncated) ---")
print("\n".join(output.splitlines()[:15]))
print("--- end ---")
def run_remote_probe(target: str) -> None:
import requests
print("\n[+] Remote debug indicator probe (best-effort)")
print(f" Target: {target}")
# This app does not easily throw exceptions from user input, so
# absence of indicators does NOT prove debug is off.
probe_urls = [
f"{target.rstrip('/')}/__nonexistent__",
]
detected = False
for url in probe_urls:
try:
resp = requests.get(url, timeout=10)
if "Werkzeug Debugger" in resp.text or "Traceback" in resp.text:
detected = True
print(" ✅ Debug indicators found")
break
except Exception as e:
print(f" ⚠️ Probe failed: {e}")
if not detected:
print(" ⚠️ No debug indicators found (this is not definitive)")
def main():
parser = argparse.ArgumentParser(description='Unfurl debug mode PoC (corrected)')
parser.add_argument('--spawn', action='store_true', help='Run local spawn check (default)')
parser.add_argument('--target', help='Target Unfurl URL for remote probe')
args = parser.parse_args()
if args.target:
run_remote_probe(args.target)
else:
run_spawn_check()
if __name__ == '__main__':
main()
Impact
If the service is exposed beyond localhost (bound to 0.0.0.0 or reverse-proxied), an attacker can access the Werkzeug debugger. This can disclose sensitive information and may allow remote code execution if a debugger PIN is obtained. At minimum, stack traces and environment details are exposed on errors.
Affected versions
Security releases
Kodem intelligence
Severity tells you how bad this could be in the worst case. It does not tell you whether you are exposed. Exploitability and impact are functions of runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A vulnerable package can sit in your dependency tree and never run.
Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter. Kodem's runtime-powered SCA identifies whether this CVE is reachable in your applications.
Already deployed Kodem?
See it in your environmentNew to Kodem? Get a demo →Remediation advice
Kodem Kai can prioritize this vulnerability in your dependency tree and generate a fix recommendation.
Frequently Asked Questions
- What is GHSA-VG9H-JX4V-CWX2? GHSA-VG9H-JX4V-CWX2 is a critical-severity security vulnerability in dfir-unfurl (pip), affecting versions <= 20250810. No fixed version is listed yet.
- Which versions of dfir-unfurl are affected by GHSA-VG9H-JX4V-CWX2? dfir-unfurl (pip) versions <= 20250810 is affected.
- Is there a fix for GHSA-VG9H-JX4V-CWX2? No fixed version is listed for GHSA-VG9H-JX4V-CWX2 yet. Monitor the advisory for updates and apply mitigations in the interim.
- Is GHSA-VG9H-JX4V-CWX2 exploitable, and should I be worried? Whether GHSA-VG9H-JX4V-CWX2 is exploitable in your environment depends on whether the vulnerable code is present and reachable. A CVSS score is a worst-case rating; it does not account for your specific deployment, configuration, or usage patterns. Kodem, an Intelligent Application Security platform, uses runtime intelligence to show which vulnerabilities actually execute in production, so you can focus on the ones that represent real risk. Get a demo
- What actually determines whether GHSA-VG9H-JX4V-CWX2 is exploitable, and how bad it is? Exploitability and impact are not fixed properties of a CVE. They depend on runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A high CVSS score on a dependency that never runs is not the same as real risk. Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter.