Threatswarm

27 scope-enforced AI agents that run the full pentest kill-cha...

Scam-checking just got a lot easier: Malwarebytes is now in Claude

For years, Malwarebytes has protected people by going where they are, and where people are today is increasingly within AI tools. As these chatbots tackle more everyday questions—like what to wear for an interview, how to replace a pendant light in the home, and where to eat during upcoming...

Claude Mythos Has Found 271 Zero-Days in Firefox

That's a lot. No, it's an extraordinary number: Since February, the Firefox team has been working around the clock using frontier AI models to find and fix latent security vulnerabilities in the browser. We wrote previously about our collaboration with Anthropic to scan Firefox with Opus 4.6, whi...

PT-2026-36109

Name of the Vulnerable Software and Affected Versions Claude SDK for TypeScript versions 0.79.0 through 0.91.0 Description The BetaLocalFilesystemMemoryTool creates memory files and directories using Node.js default modes 0o666 for files and 0o777 for directories. This results in files being...

CVE-2026-7235

A security vulnerability has been detected in ErlichLiu claude-agent-sdk-master up to b185aa7ff0d864581257008077b4010fca1747bf. Affected by this vulnerability is an unknown functionality of the file app/api/agent-output/route.ts. The manipulation of the argument outputFile leads to path traversal...

CVE-2026-7235 ErlichLiu claude-agent-sdk-master route.ts path traversal

A security vulnerability has been detected in ErlichLiu claude-agent-sdk-master up to b185aa7ff0d864581257008077b4010fca1747bf. Affected by this vulnerability is an unknown functionality of the file app/api/agent-output/route.ts. The manipulation of the argument outputFile leads to path traversal...

EUVD-2026-26007

A security vulnerability has been detected in ErlichLiu claude-agent-sdk-master up to b185aa7ff0d864581257008077b4010fca1747bf. Affected by this vulnerability is an unknown functionality of the file app/api/agent-output/route.ts. The manipulation of the argument outputFile leads to path traversal...

CVE-2026-7235 ErlichLiu claude-agent-sdk-master route.ts path traversal

A security vulnerability has been detected in ErlichLiu claude-agent-sdk-master up to b185aa7ff0d864581257008077b4010fca1747bf. Affected by this vulnerability is an unknown functionality of the file app/api/agent-output/route.ts. The manipulation of the argument outputFile leads to path traversal...

CVE-2026-7235

CVE-2026-7235 affects the ErlichLiu claude-agent-sdk-master project (up to commit b185aa7ff0d864581257008077b4010fca1747bf). The vulnerability is in app/api/agent-output/route.ts where manipulation of the outputFile argument leads to a path traversal. The issue could be remotely triggered and has...

CVE-2026-7235

A security vulnerability has been detected in ErlichLiu claude-agent-sdk-master up to b185aa7ff0d864581257008077b4010fca1747bf. Affected by this vulnerability is an unknown functionality of the file app/api/agent-output/route.ts. The manipulation of the argument outputFile leads to path traversal...

CVE-2026-7216

A weakness has been identified in donchelo processing-claude-mcp-bridge up to e017b20a4b592a45531a6392f494007f04e661bd. Impacted is an unknown function of the file processingserver.py of the component createsketch Tool. This manipulation of the argument sketchname causes path traversal. Remote...

CVE-2026-7216 donchelo processing-claude-mcp-bridge create_sketch Tool processing_server.py path traversal

A weakness has been identified in donchelo processing-claude-mcp-bridge up to e017b20a4b592a45531a6392f494007f04e661bd. Impacted is an unknown function of the file processingserver.py of the component createsketch Tool. This manipulation of the argument sketchname causes path traversal. Remote...

EUVD-2026-25972

A weakness has been identified in donchelo processing-claude-mcp-bridge up to e017b20a4b592a45531a6392f494007f04e661bd. Impacted is an unknown function of the file processingserver.py of the component createsketch Tool. This manipulation of the argument sketchname causes path traversal. Remote...

CVE-2026-7216

The CVE-2026-7216 entry describes a weakness in donchelo processing-claude-mcp-bridge up to e017b20a4b592a45531a6392f494007f04e661bd. The vulnerable component is the create_sketch Tool, specifically the processing_server.py function handling the sketch_name argument. This input manipulation enabl...

Claude Agent SDK Master 路径遍历漏洞

Claude Agent SDK Master is a progressive learning tutorial for Claude Agent SDK developed by Erlich. Claude Agent SDK Master has a path traversal vulnerability; this vulnerability stems from the outputFile parameter in the app/api/agent-output/route.ts file, which allows for path traversal,...

Processing-Claude MCP Bridge 路径遍历漏洞

Processing-Claude MCP Bridge is a bridge tool developed by Mariano, allowing for control of Processing applications via natural language. Processing-Claude MCP Bridge contains a path traversal vulnerability, which stems from the sketchname parameter in the processingserver.py file within the...

PT-2026-35681

A security vulnerability has been detected in ErlichLiu claude-agent-sdk-master up to b185aa7ff0d864581257008077b4010fca1747bf. Affected by this vulnerability is an unknown functionality of the file app/api/agent-output/route.ts. The manipulation of the argument outputFile leads to path traversal...

A week in security (April 20 – April 26)

Last week on Malwarebytes Labs: Medical data of 500,000 UK volunteers listed for sale on Alibaba How cyberattacks on companies affect everyone Apple fixes iOS bug that kept deleted notifications, including chat previews Roblox clamps down on chats and age checks as legal pressure builds Malicious...

Jailbreaking Frontier Foundation Models through Intention Deception

Large vision-language models exhibit remarkable capability but remain highly susceptible to jailbreaking. Existing safety training approaches aim to have the model learn a refusal boundary between safe and unsafe, based on the user's intent. It has been found that this binary training regime ofte...

GHSA-Q5HJ-MXQH-VV77 Claude Code: Trust Dialog Bypass via Git Worktree Spoofing Allows Arbitrary Code Execution

Claude Code used the git worktree commondir file when determining folder trust but did not validate its contents. By crafting a repository with a commondir file pointing to a path the victim had previously trusted, an attacker could bypass the trust dialog and immediately execute malicious hooks...