26672 matches found
How MCP Servers Can Expose Enterprise Secrets
MCP servers can expose enterprise secrets through plaintext configuration files, over-permissioned access and prompt injection, often before security teams even know the server is running. As more organizations adopt AI agents into their systems, that exposure can silently become a major gap in M...
BIT-NODE-MIN-2026-56850
A flaw in Node.js HTTPS Agent connection reuse can cause PFX object-array key collisions, allowing mutual TLS mTLS client identities to be reused across requests configured with different client certificates. This vulnerability affects Node.js 26.x, 24.x, and 22.x...
BIT-JENKINS-2026-70426
In Remoting 3384.v60d89463d9e0 and earlier, except 3355.3357.v931d3c992987, included in Jenkins 2.575 and earlier, LTS 2.568.1 and earlier, the JEP-200 class filter is not applied to classes resolved via a fallback path in the Remoting deserialization implementation, allowing agent processes, cod...
BIT-CONSUL-2026-19113 Unauthenticated denial of service via unbounded request body processing
Consul Community Edition and Consul Enterprise 1.3.0 through 2.0.2 are vulnerable to an unauthenticated denial of service in several agent HTTP API endpoints. A remote caller could cause the agent to consume substantial memory before the request was rejected. This vulnerability, CVE-2026-19113, i...
Malicious Package
Overview @ai-vertical/ai-agent is a malicious package. This package contains malicious code, and its content was removed from the official package manager. While this package might be attempting to impersonate a valid organization, there is no connection between that organization and this package...
MAL-2026-14070 Malicious code in @ai-vertical/ai-agent (npm)
--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector f3daa0f3fabdeebe3db0146233d623d65ae16aea601e54fd6519a160db5af8ee On require/import, index.js fetches a hardcoded GitHub Gist gist.github.com/aiverticalsolutions/7676451d2f972137d2482c4a8937a77a via the api.github.c...
K000162940: NodeJS vulnerability CVE-2026-56850
Security Advisory Description A flaw in Node.js HTTPS Agent connection reuse can cause PFX object-array key collisions, allowing mutual TLS mTLS client identities to be reused across requests configured with different client certificates. This vulnerability affects Node.js 26.x, 24.x, and 22.x...
Security of Foundation-Model-Powered Embodied Agents: Attack Surfaces, Attacks, Defenses, and Evaluation
Foundation models are increasingly used for perception, reasoning, planning, and action generation in embodied agents, creating security risks that can propagate from digital inputs to physical behavior. Existing surveys often organize threats by mechanisms such as jailbreaks, prompt injection,...
PT-2026-73342
In Remoting 3384.v60d89463d9e0 and earlier, except 3355.3357.v931d3c992987, included in Jenkins 2.575 and earlier, LTS 2.568.1 and earlier, the JEP-200 class filter is not applied to classes resolved via a fallback path in the Remoting deserialization implementation, allowing agent processes, cod...
PT-2026-73343
Jenkins 2.575 and earlier, LTS 2.568.1 and earlier does not safely handle symbolic links with effectively empty names during the extraction of .tar and .tar.gz archives, allowing attackers able to control agent processes to provide crafted archives to the controller to write files to arbitrary...
Proof-Of-Execution Memory: Defending LLM Agents against Forged-Reasoning Attacks by Verifying What Actually Happened
LLM agents are stateless and rely on external memory to carry context between steps. Because agents treat that memory as trustworthy, an adversary who can write to it can steer their behavior. The FARMA attack does this with no malicious command: it inserts fabricated entries into the agent's...
UBUNTU-CVE-2026-74270
In the Linux kernel, the following vulnerability has been resolved: handshake: Require admin permission for DONE command ACCEPT and DONE are the two downcalls of the handshake genl family, both intended for use by the trusted handshake agent tlshd. ACCEPT already requires GENLADMINPERM; DONE has ...
GPEWebDefender
GPEWebDefender A web-attack monitor. It sits next to ngin...
CVE-2026-24882 vulnerabilities
Vulnerabilities for packages: gnupg...
GHSA-M6X2-4HHH-669J vulnerabilities
Vulnerabilities for packages: gnupg...
CVE-2026-24882 vulnerabilities
Vulnerabilities for packages: gnupg...
CVE-2026-57062 vulnerabilities
Vulnerabilities for packages: gnupg...
Beyond Direct Access: Resource Hijacking in LLM Agents
Large language model agents are increasingly connected to high-value resources such as computing infrastructure, credentials, usage budgets, identities, private knowledge, communication channels, and organizational workflows. Existing agent security research mainly studies attacks on instructions...
Hierarchical Agentic Incident Response with Digital-Twin-Validated Attack Inference
Network incident response remains slow and labor-intensive as the defender must infer multi-stage attacks from partial observations and translate recovery decisions into reliable system commands. Decision-theoretic planners provide principled optimization but typically rely on abstract states and...