Loading vulnerability…
CVE-2022-35411 is an unsafe deserialization vulnerability in rpc.py through version 0.6.0. Although JSON is the default RPC data format, the framework accepts a client-controlled serializer selection through an HTTP header. An unauthenticated client can select Python pickle processing, causing attacker-supplied serialized data to be passed to unpickle. Because Python pickle deserialization can invoke attacker-controlled code paths, a crafted payload can result in remote code execution.
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.
2 valid exploits after Mallory filtered fakes, detection scripts, and README-only repos (1 hidden).
Repository contains a single Python exploit script (`rpc-exploit.py`) and a minimal README. The exploit targets rpc.py versions v0.4.2–v0.6.0 (CVE-2022-35411) and achieves unauthenticated remote code execution by sending an HTTP POST request to `http://<host>/sayhi` with header `serializer: pickle` and a malicious pickle payload. The payload is generated by defining a class with a `__reduce__` method that returns `os.system` and an attacker-supplied command, which executes upon server-side unpickling. The script’s `main()` includes a hardcoded reverse-shell command using a FIFO at `/tmp/f`, `/bin/sh -i`, and `nc` to connect back to `192.168.45.196:9091`. Overall purpose: demonstrate/weaponize unsafe pickle deserialization in rpc.py’s serializer handling to run arbitrary OS commands on the server.
This repository contains an operational exploit for CVE-2022-35411, targeting rpc.py versions 0.4.2 through 0.6.0. The exploit leverages unsafe deserialization of Python pickle objects in the rpc.py server, allowing unauthenticated remote code execution via crafted HTTP POST requests. The main exploit script, 'rpc-rce.py', is a modular Python tool that accepts command-line arguments for the attacker's host and port, the target endpoint, and a dry-run mode. By default, it sends a test payload (curl command) and, unless in dry-run mode, delivers a reverse shell payload to the attacker's listener. The README provides detailed usage instructions, requirements, and ethical guidance. The exploit is not part of a framework and is self-contained, with the main entry point being 'rpc-rce.py'. The attack vector is network-based, exploiting an HTTP endpoint exposed by the vulnerable rpc.py server. The code is clean, well-documented, and suitable for both proof-of-concept and practical exploitation in authorized testing scenarios.
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.
3 sources tracked across advisories, community write-ups, and news. New activity surfaces here as Mallory finds it.
A critical unauthenticated remote-code-execution vulnerability in rpc.py through version 0.6.0. An attacker can submit the serializer: pickle HTTP header to trigger unsafe pickle deserialization despite JSON being the default format.
A Python RPC-framework vulnerability in which default-enabled pickle functionality permits clients to unpickle untrusted data.
A vulnerability in a Python-based RPC framework arising from default-enabled pickle functionality that permits unpickling untrusted data.
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.