TL;DR
- Cursor says a compromised key at AI provider Fireworks led to malicious Claude Code extensions, and the breach is still the clearest supply-chain case this month Dark Reading.
- Google describes attackers running autonomous multi-agent tools, but the Java and "first fully autonomous attack" headlines are weak leads The Hacker News dev.ua Cyber Magazine.
- Supergateway and vLLM both have new CVEs where an unauthenticated or hostile client can reach the agent or model backend CVE-2026-132121 CVE-2026-130686 CVE-2026-130685.
The week in agent security
This week the risk sits at the edges of the agent, not inside the model. A stolen vendor key reached Cursor data and credentials Dark Reading. A researcher group took over a popular model account on Hugging Face because the platform does not verify ownership BleepingComputer. In both cases, trust in a third party was the weak point.
The same pattern shows up in serving software. Supergateway lets a web page talk to a local MCP process CVE-2026-132121. Two vLLM bugs let remote clients exhaust resources or crash the engine CVE-2026-130686 CVE-2026-130685. These are classic web bugs in new places. In our view, many AI tools were built for a trusted laptop, then exposed to networks and browsers.
On the attacker side, Google says threat actors now use autonomous frameworks such as ARTEX, CAIRN, CyberStrikeAI and PentAGI to build and run exploits from patches The Hacker News. Reports on ransomware and a phishing kit add that AI mostly gives scale and better lures Dark Reading Dark Reading. The defence answer is the same as always: limit what a stolen key or a steered agent can reach. Simon Willison's work on external sandboxes follows that idea Simon Willison.
Top stories
Supply chain
Fireworks key leads to malicious Claude Code extensions
Cursor traced malicious Claude Code extensions to a compromised key issued to Fireworks. Threat actors used it to reach Cursor data and credentials Dark Reading. The excerpt does not say what the extensions did or how many people installed them. Anyone who installed them should treat local credentials as exposed.
Threat level: High. Real compromise with a confirmed path, but payload details are missing.
Tool security
Supergateway cross-site WebSocket hijacking
CVE-2026-132121 affects Supergateway 2.4.0 through 4.2.2 in stdio-to-WebSocket mode. It does not check the Origin header, so a malicious page can open the /message WebSocket from the victim's browser CVE-2026-132121. In our view, that could let a web page drive a local MCP server and its tools.
Threat level: High. Public details on a real MCP bridge; upgrade or restrict access.
Model security
Hugging Face account takeover of Kokoro
Researchers took control of the account for Kokoro, a text-to-speech model used in projects from Google, NVIDIA, AMD, Intel and Amazon Web Services. The cause was missing ownership verification BleepingComputer. Anyone who pulls models by name could be served changed files.
Threat level: Medium. Done by researchers, but the weakness is real and the reach is wide.
vLLM resource exhaustion and engine crash
CVE-2026-130686 lets remote clients exhaust frontend resources with many image URLs in one request CVE-2026-130686. CVE-2026-130685 lets unauthenticated attackers crash EngineCore with out-of-range token ids when --skip-tokenizer-init is used CVE-2026-130685. Both are fixed in 0.33.0 per the excerpts.
Threat level: Medium. Denial of service only, but it affects shared inference hosts.
Expert take: threats and trends
Compared with last week, the Fireworks breach continued, and its main facts are unchanged Dark Reading. Supergateway continued as a story, now as a known CVE rather than a new find CVE-2026-132121. Google's autonomous agent report continued, and no new frameworks were named The Hacker News. Uber's victim count did not appear this week.
New this week: model-hosting trust (Hugging Face) and inference server bugs (vLLM). Attackers are going after the supply chain around agents, because it is easier than attacking the model.
Over-hyped: the "AI found a critical Java flaw threatening 90% of top companies" headlines. These are search leads with no detail, so we do not rate them dev.ua Anadolu Ajansı. Same for "first fully autonomous AI attacks" Cyber Magazine and the vendor pentest pieces, which are marketing-adjacent The Hacker News The Hacker News. The prompt injection explainer is background only qz.com. We skip the currency items as noise link link.
Watch next week: Details of what the malicious extensions did, and any exploit code for the Supergateway and vLLM CVEs.
Defender checklist
- Rotate keys shared with AI vendors, limit their scope, and set short expiry Dark Reading.
- Find and remove the malicious Claude Code extensions, and treat credentials on those machines as exposed Dark Reading.
- Upgrade or firewall Supergateway, and check for Origin validation on any local MCP bridge CVE-2026-132121.
- Upgrade vLLM to 0.33.0 or later, and put authentication and request limits in front of it CVE-2026-130686 CVE-2026-130685.
- Pin Hugging Face models by commit hash and mirror trusted copies BleepingComputer.
- Run coding agents in an external sandbox without long-lived credentials Simon Willison.
Detection ideas
- Add a scanner rule for MCP bridge configs that bind to all interfaces or lack Origin checks, with a Supergateway sample CVE-2026-132121.
- Flag model references that use a mutable name instead of a pinned revision, mapped to supply chain risks in OWASP LLM Top 10 BleepingComputer.
- Add a rule for extension or skill manifests that read credential stores, based on the Fireworks case Dark Reading.
- Flag vLLM launch commands using
--skip-tokenizer-initon exposed ports CVE-2026-130685.
Reading list
This week's full list of leads. Check the original sources before you act.
Independent
- External sandboxes for coding agents (2026-11-17, Simon Willison)
- datasette-llm 0.1a11 (2026-11-14, Simon Willison)
- AI Doesn’t Have to Go Rogue to Break the World (2026-11-13, Bruce Schneier)
Advisory
- CVE-2026-132121: Supergateway 2.4.0 through 4.2.2 contains a cross-site WebSocket hijacking vulnerability in... (2026-11-18, NVD CVEs)
- CVE-2026-130686: vLLM from 0.6.0 before 0.33.0 lacks a request-wide media budget, allowing remote inference... (2026-11-18, NVD CVEs)
- CVE-2026-130685: vLLM from 0.28.0 before 0.33.0 contains an improper input validation vulnerability that... (2026-11-18, NVD CVEs)
News
- AI Pentesting Checklist for Web Apps: 15 Questions to Ask Before Choosing a Vendor (2026-11-18, The Hacker News)
- Quality Isn’t The Main Worry About AI Pentesting - Data Privacy and Production Safety Are (2026-11-18, The Hacker News)
- 'TheMailer' Phishing Kit Puts AI in the Driver's Seat (2026-11-17, Dark Reading)
- Cursor Traces Malicious Claude Code Extensions to Third-Party Breach (2026-11-17, Dark Reading)
- Researchers hijack Hugging Face account for popular Kokoro AI model (2026-11-17, BleepingComputer)
- Ransomware Attacks Increase Behind AI, but Payments Decline (2026-11-16, Dark Reading)
- Google Finds Hackers Using AI Agents to Build and Run Exploits From Software Patches (2026-11-16, The Hacker News)
Discovery
- OpenAI's AI model found a critical vulnerability in Java that threatens 90% of top companies (2026-11-18, dev.ua)
- AI model uncovers critical security vulnerability in Java: Report (2026-11-18, Anadolu Ajansı)
- Prompt injection attacks explained: how hackers hijack AI (2026-11-17, qz.com)
- Google Spots Fully Autonomous AI Attacks For the First Time (2026-11-17, Cyber Magazine)
- How do AI vulnerability and RBA limits affect the Australian Dollar? – MUFG (2026-11-14, link)
- Australian Dollar: AI vulnerability and RBA limits – MUFG (2026-11-14, link)
END OF TRANSMISSION // Verify the signal. Defend the perimeter.