Loading vulnerability…
CVE-2026-33033 is a resource-consumption vulnerability in Django's MultiPartParser. Crafted multipart file uploads using Content-Transfer-Encoding: base64 with excessive whitespace may trigger repeated memory copying, degrading application performance and potentially causing denial of service. Affected releases include Django 6.0 before 6.0.4, 5.2 before 5.2.13, and 4.2 before 4.2.30. Earlier unsupported series, including 5.0.x, 4.1.x, and 3.2.x, were not evaluated and may also be affected.
Mallory correlates every CVE against your assets, your vendors, and active adversary campaigns. Know which vulnerabilities matter for you, not just which ones are loud.
What it means. What to do now. Patch path, mitigations, and the assume-compromise checklist.
What an attacker gets, and what they’ve been doing with it.
If you can’t patch tonight, do this now.
Patch, then assume compromise.
1 valid exploit after Mallory filtered fakes, detection scripts, and README-only repos.
This repository is a real Python proof-of-concept for CVE-2026-33033, a Django denial-of-service issue in MultiPartParser when handling multipart file uploads marked with Content-Transfer-Encoding: base64 and containing excessive whitespace. The main exploit is exploit.py, which builds two multipart bodies of equal size: a benign base64 payload for baseline timing and a malicious payload composed of 'AAA', a large run of spaces, and a trailing 'A'. It then POSTs these bodies to a target upload endpoint and measures response times across multiple rounds to demonstrate CPU amplification and worker exhaustion. The exploit is operational rather than just demonstrative because it actively sends attack traffic and reports whether the target appears vulnerable based on elapsed time. It does not provide code execution or persistence; its sole capability is application-layer DoS through expensive request parsing. Repository structure is simple: exploit.py is the attacker script; requirements.txt pins Django 5.0.x and uWSGI; the victim/ directory contains a minimal Django application used to reproduce the issue locally. victim/victim/urls.py exposes /upload and /health. The upload handler accesses request.FILES, ensuring multipart parsing occurs. settings.py intentionally keeps default upload-related limits and includes CSRF middleware to show the vulnerable parsing path can occur before view logic. uwsgi.ini configures a 4-process HTTP listener on 0.0.0.0:8000, illustrating how a small number of concurrent malicious requests can tie up all workers. Notable observables include the default target URL http://127.0.0.1:8000/upload, derived health URL http://127.0.0.1:8000/health, the demo server listener 0.0.0.0:8000, and the crafted multipart boundary ----CVE2026-33033. Overall, the repository’s purpose is to reproduce and benchmark the DoS condition against vulnerable Django versions, not to gain shell access or deploy a secondary payload.
Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.
Vendor-confirmed product mapping. Mallory continuously reconciles this list against your asset inventory.
4 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.
Query your assets running an affected version, and investigate the blast radius.
Every observed campaign linking this CVE to a named adversary.
Malware families riding this exploit, with evidence and IOCs.
YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.
Cross-references every affected SKU, including bundled OEM variants.
Community discussion across Reddit, Mastodon, and other social sources.