Engineering a security harness for AI coding agents | Blog | Endor Labs
A capable frontier model isn’t a controlled system. Here are seven questions security teams should answer before they trust AI coding agents with consequential work.
Vendor news
Funding rounds, leadership changes, incidents, new capabilities and announcements across the vendors in this catalog — collected automatically from vendors' own public feeds. Product release versions are on their own page, so the landing feed here never reads as one vendor's changelog.Where this comes fromEvery item links to the vendor's own published source: a GitHub release feed, a changelog, a status page, an SEC filing or a company blog. Nothing here is written or editorially reviewed by 0-Doubt, and nothing is a recommendation — it is machine-collected public record, labelled the same way as all other baseline content on the platform. Coverage is uneven by design: a vendor that publishes no public feed will not appear, and that absence is not a judgement about the vendor.
608 matching stories — announcement, in Software Supply Chain Security — showing up to 3 per vendorPer-vendor capWhile browsing, no single vendor may fill more than three slots on any one page, so a vendor with a chatty release cadence cannot buy the page you are looking at. That means a page can render fewer stories than the count above. Searching a headline or naming a vendor turns the cap off, because then you have asked to see everything that matches.
Malicious npm campaign targets developers integrating Twilio
The package poses as a security tool targeting developers looking to implement Internet-based apps with telecom networks.
Malicious Package Detection: Why Databases Miss Attacks and Where Detection Has to Fire
Malicious package detection is the practice of identifying open source packages that were published or modified to attack the systems that install them... The post Malicious Package Detection: Why Databases Miss Attacks and Where Detection Has to Fire appeared first on Cycode.
Growing the WIN AI Ecosystem with Agent Integrations
WINning AI with AI: How the Wiz MCP for WIN partners accelerates a connected ecosystem
Software Development Security: A Requirements Checklist With 12 Lines You Can Verify
Most software development security requirements describe a desired state, which means nobody can pass or fail them. Here are 12 written as checks, each with the artifact that proves it. The post Software Development Security: A Requirements Checklist With 12 Lines You Can Verify appeared first on Xygeni AI-Powered AppSec Platform.
So I asked my agent instead…
Ask your Evo tenant about models, MCP servers, and skills across your AI estate, from the client you already work in.
The real and hidden costs of AI security programs
As organizations race to adopt AI across software development and security, the pressure to move fast is colliding with the mandate to stay economical, repeatable, and defensible. The biggest risk isn’t adopting too slowly. It’s scaling without understanding the true cost of AI. Prefer to listen or watch? Check out our on-demand webinar on this […] The post The real and hidden costs of AI security programs appeared first on Black Duck.
Dragos Completes NetRise and runZero Acquisitions Following Accenture Deal
The transaction is part of the $4.1 billion deal in which Accenture acquired a majority stake in Dragos in an OT cybersecurity push. The post Dragos Completes NetRise and runZero Acquisitions Following Accenture Deal appeared first on SecurityWeek.
LLM Development Risks: Security Threats and Mitigation Strategies | Blog | Endor Labs
Learn the top llm development risks, from prompt injection to data leakage, plus practical mitigation strategies for secure AI apps
Vulnerability Blast Radius: How to Measure and Reduce Impact | Blog | Endor Labs
Learn how to assess vulnerability blast radius, understand BlastRADIUS, and reduce risk with reachability, segmentation, and fixes
AI Agent Harness: The Layer That Decides Whether AI Security Works
In March 2026, LangChain took its coding agent from Top 30 to Top 5 on Terminal Bench 2.0 without touching the model or fine-tuning a single weight... The post AI Agent Harness: The Layer That Decides Whether AI Security Works appeared first on Cycode.
Xygeni Malicious Code Digest 88
81 malicious package versions confirmed this week: one npm package published 39 times in 20 minutes, plus an active Strapi plugin campaign. The post Xygeni Malicious Code Digest 88 appeared first on Xygeni AI-Powered AppSec Platform.
AI Red Team Tools Test What Your Model Says. They Don’t Test What Your Agent Does Next
Uncover the details of The 24712-pl Campaign and the zero-dependency packages exploited during a security incident. The post AI Red Team Tools Test What Your Model Says. They Don’t Test What Your Agent Does Next appeared first on Xygeni AI-Powered AppSec Platform.
Agentic AI permissions: A core problem — but not the only one
Controlling coding agent overpermissioning is key to security. But recent frontier AI incidents show that problems don’t stop there.
Exploring the new AWS Sign Up experience
This post will explore what this new concept does, how it works with the new Account Access capability, and why a strong security posture still requires upgrading out of the sandbox.
Building an AI Detection Engine That Understands Agent Intent
Analyzing model input and output logs in an AI-native detection pipeline to understand and uncover malicious AI agent behavior
13 Best DAST Tools and Scanners in 2026
This guide compares the best DAST tools on the market and explains what each one actually does well once it runs against your applications... The post 13 Best DAST Tools and Scanners in 2026 appeared first on Cycode.
Memory-safe programming goes from advocacy to adoption
The coding languages shift — fueled by Microsoft, Google, and Amazon and enabled by AI coding — is driven by the quest for secure software.
JFrog Artifactory Supports LuaRocks Hosting for NGINX, OpenResty and Kong
If your NGINX/OpenResty servers or Kong gateways pull Lua modules straight from public luarocks.org, one upstream outage can stall every build and deploy that depends on a “rock”. With LuaRocks support in JFrog Artifactory, you host and proxy those modules from a private LuaRocks registry you control, using the native experience you expect, not a …
Parallels Desktop Flaw Lets Non-Admin Mac Users Gain Root, but Intel Macs Can't Install Fix
Parallels Desktop for Mac has a flaw that lets an ordinary local account run code as root, the highest level of access on a Mac, software company JFrog said this week. The attack needs code already running on the machine as a normal user, so it does not work over the network. JFrog says the fix is in Parallels Desktop 27, a version that Intel Macs cannot install. Yuval Moravchick, who leads
Why Is Detection and Response Too Slow for Machine Speed Attacks?
TLDR: Runtime security built around detecting activity, correlating signals and then responding assumes there is enough time to act after something malicious happens. Machine speed attacks challenge that assumption because attackers can move across multiple MITRE ATT&CK tactics in seconds, with each successful action creating the conditions for the next. The issue is not whether …
How Do You Build Security for AI Agents That Don’t Exist Yet?
AI agents are rapidly evolving toward more autonomy, more context, greater reach and more authority to act. We have to build security now for what’s coming next. We are making enormous bets right now about how to secure software two years from now, at exactly the moment when software is changing too quickly for anyone […] The post How Do You Build Security for AI Agents That Don’t Exist Yet? appeared first on OX Security.
OX Cloud: CNAPP Coverage Plus Runtime Security for AI Agents
Built for a new class of active, dynamic risk that traditional cloud security wasn’t built to see. TL;DR OX Cloud is the runtime pillar of the OX AINAPP platform, and our answer to a question most cloud security tools still aren’t built to ask: what is actually happening in your cloud right now, and does […] The post OX Cloud: CNAPP Coverage Plus Runtime Security for AI Agents appeared first on OX Security.
Who's governing your AI? A trust framework for enterprise agents and models
SPONSORED FEATURE: DigiCert wants to hand every agent a passport, complete with an expiry date and a named human owner
ParaShells: Parallels Desktop Turns Appliance Install Into a Root Shell
Your Mac runs a vulnerable version of Parallels Desktop. A malicious package, compromised CI job, or other unprivileged process is already running on it. No admin access. No Parallels-signed client. One appliance-install request later, attacker-controlled code runs as root. While testing Desktop 26.4.0 (build 57513) on Apple silicon, we found that an unprivileged local user …
HIPAA Enforcement Is Increasing: Why Third-Party Scripts Are the Next Big Target
A healthcare marketer adds a “simple” tracking pixel to an appointment page. Nothing exotic. Just a script to measure conversions, tune campaigns,... Read More... The post HIPAA Enforcement Is Increasing: Why Third-Party Scripts Are the Next Big Target appeared first on Source Defense.
The AI Hurricane Is Here
AI is accelerating software creation and cyberattacks alike. Leaders must secure agents and code at inception, enforce controls at runtime, and validate defenses independently.
Four Critical CVEs, the Same Trust Issue
Last week, 4 unauthenticated critical CVEs turned up in 24 hours, all sharing the same mistake: a component trusting the layer next to it Read our technical analysis of these 4 critical CVEs Within a single 24-hour window last week, four critical vulnerabilities landed across widely deployed infrastructure: two in Next.js, one in Netty, one […] The post Four Critical CVEs, the Same Trust Issue appeared first on OX Security.
OpenAI Agents Linked to RubyGems Campaign That Gained RCE on RubyDoc Servers
The "major malicious attack" that targeted RubyGems in May 2026 was the work of a swarm of OpenAI agents, according to a new report published by researchers Spencer Kitts, Thomas Larsen, and Sydney Von Arx. On May 12, Maciej Mensfeld, senior product manager for software supply chain security at Mend.io, disclosed details of a coordinated cyber attack that targeted the package manager for the
래브라도랩스, 공급망 노리는 악성 오픈소스 패키지, 도입 전 막는다
Agentic AppSec on your infrastructure with Black Duck Signal
Agentic development has enabled enterprises to accelerate the pace of innovation while freeing developers to focus on more strategic initiatives. But for many regulated organizations, the idea of sending source code and analysis prompts to vendor-managed LLM endpoints is a nonstarter. That’s why Black Duck Signal™ added support for agentic AppSec testing using your own […] The post Agentic AppSec on your infrastructure with Black Duck Signal appeared first on Black Duck.
What PSPs Aren’t Telling Merchants About Client-Side Risk
Payment service providers play a critical role in modern commerce. They help merchants accept payments, reduce infrastructure complexity, and shift sensitive parts... Read More... The post What PSPs Aren’t Telling Merchants About Client-Side Risk appeared first on Source Defense.
Detection scaled. The loop did not.
Findings got cheap. Verified closure did not. That gap is the AppSec problem.
Is prevention essentially a solved problem?
Prevention in agent-generated code is architecturally solved—but choosing controls that protect security without slowing development remains the challenge.
The vulnerability you didn’t write
Your base image might contain someone else’s problem, right up until you ship it Pulling a base image is routine. You grab it, build on top of it, ship the container, and move on. It is fast, reproducible, and usually reliable. The catch is that every base image comes with history, and that history is […] The post The vulnerability you didn’t write appeared first on Black Duck.
A Holistic Look at eSkimming: From Formjacking to Full Funnel Manipulation
How modern website attacks evolve inside the browser, and what to do about it It starts like a lot of incident reviews... Read More... The post A Holistic Look at eSkimming: From Formjacking to Full Funnel Manipulation appeared first on Source Defense.
Mend Renovate Enterprise: managing dependencies at agentic scale
How Mend Renovate Enterprise keeps dependency updates flowing at agentic scale.
Runtime Security for AWS CodeBuild-Hosted GitHub Actions Runners
Harden-Runner now secures GitHub Actions jobs running on AWS CodeBuild-hosted runners, on EC2 compute, with managed or custom images.
Agentic Networks Need an Autonomous Control Layer: Why AgentX Is Becoming Urgent
Agentic networks are moving fast. Without an autonomous control layer, so is the risk. Enterprises are rapidly discovering that monitoring AI agents is only half of the security problem. Once agents can execute commands, invoke APIs, change infrastructure and coordinate with other agents, organizations also need a way to control what happens next. That is the role of AgentX.
@7nohe/openapi-react-query-codegen Compromised Through an Exposed npm Publishing Workflow
Ten malicious versions of @7nohe/openapi-react-query-codegen were published through an exposed npm publishing workflow. IOCs and recovery steps inside.