CVE-2026-31254

The flash-attention project thru commit e724e2588cbe754beb97cf7c011b5e7e34119e62 2025-13-04 contains a code injection vulnerability CWE-94 in its training script. The script registers the Python eval function as a Hydra configuration resolver under the name eval. This allows configuration files t...

GHSA-WXXX-GVQV-XP7P LiteLLM has a sandbox escape in custom-code guardrail

Impact The POST /guardrails/testcustomcode endpoint runs user-supplied Python inside a hand-rolled sandbox. The sandbox can be escaped using bytecode-level techniques, allowing arbitrary code execution in the proxy process — which runs as root in the default Docker image. Reaching the endpoint...

Hackers Used AI to Develop First Known Zero-Day 2FA Bypass for Mass Exploitation

Google on Monday disclosed that it identified an unknown threat actor using a zero-day exploit that it said was likely developed with an artificial intelligence AI system, marking the first time the technology has been put to use in the wild in a malicious context for vulnerability discovery and...

Malicious code in xxx-bale (PyPI)

--- -= Per source details. Do not edit below this line.=- Source: kam193 1109b5dc74c94551027044e54e20f9c1c18f89d53da6af87861ba4773eae1966 The package contains code to install remotely stored malware and ensure its persistence. The code is not triggered automatically; it requires a separate trigge...

Directory Traversal

Overview python-liquid is an A Python engine for the Liquid template language. Affected versions of this package are vulnerable to Directory Traversal via the FileSystemLoader and CachingFileSystemLoader components. An attacker can access and render arbitrary files outside the intended search pat...

GHSA-8P4X-WR7X-3788 python-liquid: Absolute paths escape filesystem loader search path

Impact The built-in FileSystemLoader and CachingFileSystemLoader do not guard against reading files outside their search paths when given an absolute path to resolve. This allows malicious template authors to load and render arbitrary files via the % include % and % render % tags. Targeted files...

GHSA-9MQQ-JQXF-GRVW PraisonAI MCP `tools/call` path-traversal => RCE via Python `.pth` injection

Summary PraisonAI's MCP Model Context Protocol server praisonai mcp serve registers four file-handling tools by default — praisonai.rules.create, praisonai.rules.show, praisonai.rules.delete, and praisonai.workflow.show. Each accepts a path or filename string from MCP tools/call arguments and joi...

CLSA-2026-1778505256 python: Fix of 2 CVEs

CVE-2021-3733: fix ReDoS in urllib2 AbstractBasicAuthHandler regex; the legacy '?:.,' prefix is replaced with the upstream-3.x form '?:^|,' and the scheme charset excludes ',' to prevent quadratic backtracking on crafted WWW-Authenticate headers - CVE-2021-23336: stop accepting ';' as a default...

Malicious code in mpkg123 (PyPI)

--- -= Per source details. Do not edit below this line.=- Source: kam193 df9e0498d827adeb16ea11e4a1137133d2124f039942b776f7ac098a257cd164 If executed as a module, the obfuscated code collects and exfiltrates sensitive data, including passwords saved in a browser. --- Category: MALICIOUS - The...

Malicious code in xxoo-bale (PyPI)

--- -= Per source details. Do not edit below this line.=- Source: kam193 74ce2be8301ccea70138e307282fbf70ede26eede2a531296145f7d0da695b80 The package contains code to install remotely stored malware and ensure its persistence. The code is not triggered automatically; it requires a separate trigge...

OPENSUSE-SU-2026:20737-1 Security update for python-lxml

This update for python-lxml fixes the following issue - CVE-2026-41066: Information disclosure via untrusted XML input leading to local file read bsc1263254...

SUSE-SU-2026:21603-1 Security update for python-lxml

This update for python-lxml fixes the following issue - CVE-2026-41066: Information disclosure via untrusted XML input leading to local file read bsc1263254...

SUSE-SU-2026:21587-1 Security update for python-lxml

This update for python-lxml fixes the following issue - CVE-2026-41066: Information disclosure via untrusted XML input leading to local file read bsc1263254...

CLSA-2026-1778489354 python2: Fix of CVE-2026-6100

CVE-2026-6100: defensively null bzs-nextin on the error path of BZ2Decompdecompress to align with upstream; the UAF window does not exist in Python 2.7 nextin is reassigned at function entry, lzma/gzip are not C extensions...

Fake OpenAI Privacy Filter Repo Hits #1 on Hugging Face, Draws 244K Downloads

A malicious Hugging Face repository managed to take a spot in the platform's trending list by impersonating OpenAI's Privacy Filter open-weight model to deliver a Rust-based information stealer to Windows users. The project, named Open-OSS/privacy-filter, masqueraded as its legitimate counterpart...

Malicious code in dlocal-cli (PyPI)

--- -= Per source details. Do not edit below this line.=- Source: kam193 9cfdf8d83ac7dc528caac3292d1b02ba162629b349789149fbbfcb7094f778b0 Generic campaign for all likely research / pentests, where the amount or art of collected data raises questions about the privacy, security and ethical side. -...

PT-2026-39742

CVE-2026-20352 iOS 26.3-Research A Public Open-Source research framework with .py and .sh files created for analyzing iOS 26.3 security mechanisms. This project is designed to be advanced through the collective in... https://t.co/5O6AR6f6H7...

📄 CairoSVG Denial of Service

CairoSVG versions prior to 2.9.0 suffer from a recursive denial of service vulnerability. CVE-2026-31899: Exponential DoS via Recursive Element Amplification in CairoSVG Keywords: CVE-2026-31899, CairoSVG, exponential DoS, SVG bomb, recursive use element, denial of service, XML amplification,...

FlashAttention 安全漏洞

FlashAttention is an efficient and memory-efficient attention mechanism implementation tool open-sourced by Dao AI Lab. There is a security vulnerability in FlashAttention; this vulnerability stems from the training script registering the Python eval function as a Hydra configuration parser, whic...

OverrideFuzz: Semantic-Aware Grammar Fuzzing for Script-Runtime Vulnerabilities

Script-language runtimes such as Python, Lua, and JavaScript are widely deployed in security sensitive contexts, yet they remain difficult to test because valid inputs must satisfy syntax, dynamic type constraints, and object-level semantics. Existing grammar and reflection-based fuzzers improve...