CVE-2025-54794-Hijacking-Claude-AI-with-a-Prompt-Injection-The-Jailbreak-That-Talked-Back

🧠 CVE-2025-54794: 利用提示注入劫持Claude AI – 会回应的越狱攻击 作者：Aditya Bhatt | 进攻安全专家 | 红队操作员 | VAPT爱好者 ⚔️ 引言：当你的AI能被文字攻破时 在一个语言模型已成为我们代码、内容和认知副驾驶的时代，漏洞不再仅仅关乎端口和有效载荷。它们关乎文字 。 CVE-2025-54794 不仅仅是CVE档案中的又一个编号——它是一句宣言： “即使是最先进的AI，也能被恰到好处的低语所操纵。” 这个高严重性的提示注入漏洞针对的是Claude AI...

DEBIAN-CVE-2026-86430

league/commonmark versions before 2.9.1 contain multiple denial of service vulnerabilities in fenced code block detection, reference link label lookup, and emphasis delimiter processing that perform super-linear work on crafted input. Attackers can submit specially crafted Markdown with long...

EUVD-2026-72388

league/commonmark versions before 2.9.1 contain multiple denial of service vulnerabilities in fenced code block detection, reference link label lookup, and emphasis delimiter processing that perform super-linear work on crafted input. Attackers can submit specially crafted Markdown with long...

PT-2026-86796

league/commonmark versions before 2.9.1 contain multiple denial of service vulnerabilities in fenced code block detection, reference link label lookup, and emphasis delimiter processing that perform super-linear work on crafted input. Attackers can submit specially crafted Markdown with long...

CVE-2026-70480: Server-Side Request Forgery (SSRF)

Open WebUI is an extensible, feature-rich, and user-friendly self-hosted AI platform. From 0.6.34 until 0.11.0, Open WebUI renders vega and vega-lite fenced code blocks in chat content by building a Vega view in the viewer browser without a restricted resource loader. Any user who can place such ...

Rogue Agent Flaw Could Have Let Attackers Hijack Google Dialogflow CX Chatbots

A critical flaw in Google's Dialogflow CX could have let an attacker with edit rights on one Code Block-enabled agent compromise other Code Block-enabled agents in the same Google Cloud project. From there, they could read live conversations, steal the data users shared, and make the bots send...

CVE-2026-58402

Hugo (static site generator) from version 0.60.0 up to 0.163.3 is affected: its default code-block renderer writes the Markdown code-fence language/info-string into the code element wrapper (class="language-…", data-lang="…") without HTML escaping. A fence info-string containing a quote and a scr...

CVE-2026-53428

Summary: The CVE describes an unbounded memory allocation in the mdex/native codepath when parsing a user-supplied highlight_lines range in Markdown code blocks, enabling a denial-of-service via memory exhaustion. The affected components are mdex (v0.11.0 before 0.12.3) and mdex_native (v0.1.0 be...

CVE-2026-49493

Markdown Preview Enhanced before 0.8.28 parses Bitfield fenced code blocks with interpretJS, which evaluates the block content as code via vm.runInNewContext, allowing arbitrary code execution. A crafted markdown document containing a malicious bitfield code block executes attacker-controlled cod...

CVE-2026-49493

Markdown Preview Enhanced before 0.8.28 parses Bitfield fenced code blocks with interpretJS, which evaluates the block content as code via vm.runInNewContext, allowing arbitrary code execution. A crafted markdown document containing a malicious bitfield code block executes attacker-controlled cod...

CVE-2026-49493: Improper Control of Generation of Code

Markdown Preview Enhanced before 0.8.28 parses Bitfield fenced code blocks with interpretJS, which evaluates the block content as code via vm.runInNewContext, allowing arbitrary code execution. A crafted markdown document containing a malicious bitfield code block executes attacker-controlled cod...

CVE-2026-39965

Summary: CVE-2026-39965 affects TypeBot (versions ≤ 3.15.2). The HTTP Request and Code blocks validate the initial URL but the HTTP clients (ky and fetch) do not re-validate redirect destinations on 302 responses, enabling an authenticated user to point a block to an attacker-controlled server th...

Linux Distros Unpatched Vulnerability : CVE-2020-37121

The Linux/Unix host has one or more packages installed that are impacted by a vulnerability without a vendor supplied patch available. - CODE::BLOCKS 16.01 contains a buffer overflow vulnerability that allows attackers to execute arbitrary code by overwriting Structured Exception Handler with...

DEBIAN-CVE-2020-37121

CODE::BLOCKS 16.01 contains a buffer overflow vulnerability that allows attackers to execute arbitrary code by overwriting Structured Exception Handler with crafted Unicode characters. Attackers can create a malicious M3U playlist file with 536 bytes of buffer and shellcode to trigger remote code...

UBUNTU-CVE-2020-37121

CODE::BLOCKS 16.01 contains a buffer overflow vulnerability that allows attackers to execute arbitrary code by overwriting Structured Exception Handler with crafted Unicode characters. Attackers can create a malicious text file with 1982 bytes of buffer and shellcode to trigger remote code...

CVE-2020-37121

CODE::BLOCKS 16.01 contains a buffer overflow vulnerability that allows attackers to execute arbitrary code by overwriting Structured Exception Handler with crafted Unicode characters. Attackers can create a malicious M3U playlist file with 536 bytes of buffer and shellcode to trigger remote code...

CVE-2020-37121

Removed by vendor...

CVE-2020-37121

CODE::BLOCKS 16.01 has a buffer overflow vulnerability that enables remote code execution by overwriting the Structured Exception Handler with crafted Unicode in a malicious M3U playlist (~536 bytes). Affected component is the SEH handler logic within CODE::BLOCKS 16.01; exploitation details spec...

EUVD-2020-31020

CODE::BLOCKS 16.01 contains a buffer overflow vulnerability that allows attackers to execute arbitrary code by overwriting Structured Exception Handler with crafted Unicode characters. Attackers can create a malicious M3U playlist file with 536 bytes of buffer and shellcode to trigger remote code...

Code::Blocks 安全漏洞

Code::Blocks is an integrated development environment developed by Code::Blocks Open Source. Version 16.01 of Code::Blocks contains a security vulnerability; this vulnerability stems from a buffer overflow, which could allow attackers to execute arbitrary code...