# Aikido Security's Blog

Discover today's best security practices and the latest trends that your software company should be aware of. Stay ahead of the game and read Aikido's industry-leading blog today.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## AI incidents of 2026: Rogue agents, stolen keys, and other familiar stories

DevFeed: [AI incidents of 2026: Rogue agents, stolen keys, and other familiar stories](<https://devfeed.tech/articles/ai-incidents-of-2026-rogue-agents-stolen-keys-and-other-familiar-stories-143870.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/ai-incidents-root-causes-patterns>)

Author: Oscar Sanchez Jr.

Published: 2026-11-19T16:04:00Z

Content type: news

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [Security](<https://devfeed.tech/topics/security.md>), [LLM security](<https://devfeed.tech/topics/llm-security.md>), [Secret Scanning](<https://devfeed.tech/topics/secret-scanning.md>), [Secrets Management](<https://devfeed.tech/topics/secrets-management.md>), [Threat monitoring](<https://devfeed.tech/topics/threat-monitoring.md>), [account takeover](<https://devfeed.tech/topics/account-takeover.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [exposed](<https://devfeed.tech/tags/exposed.md>), [incidents](<https://devfeed.tech/tags/incidents.md>), [isolation](<https://devfeed.tech/tags/isolation.md>), [keys](<https://devfeed.tech/tags/keys.md>), [news](<https://devfeed.tech/tags/news.md>), [secrets](<https://devfeed.tech/tags/secrets.md>), [security](<https://devfeed.tech/tags/security.md>), [security-incidents](<https://devfeed.tech/tags/security-incidents.md>)

### AI overview

A review of 50 publicly reported AI-related security incidents through November 19, 2026 found that familiar weaknesses dominated: exposed secrets, excessive permissions, weak isolation and compromised vendors. AI agents sometimes accelerated attacks or acted beyond intended boundaries, but only three incidents involved a reported previously unknown vulnerability. Of 44 incidents with a stated cause, 35 had multiple failure points; six reports gave no specific cause and 17 lacked at least one detail. The review recommends limiting agents' access, protecting credentials, enforcing boundaries and requiring approval for irreversible actions. Its counts are based on disclosed cases, categories can overlap, and the sample is not a complete picture of incidents that year.

### Source excerpt

We reviewed 50 AI-related security incidents from 2026. Most came down to exposed secrets, over-scoped access, weak isolation and compromised vendors. Category: News

## Aikido launches agentic pentesting for iOS apps

DevFeed: [Aikido launches agentic pentesting for iOS apps](<https://devfeed.tech/articles/aikido-launches-agentic-pentesting-for-ios-apps-143871.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/aikido-launches-agentic-pentesting-for-ios-apps>)

Author: Mario Popescu

Published: 2026-11-19T12:00:00Z

Content type: release

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [iOS](<https://devfeed.tech/topics/ios.md>), [Android Security](<https://devfeed.tech/topics/android-security.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-pentesting](<https://devfeed.tech/tags/agentic-pentesting.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [ai-pentesting](<https://devfeed.tech/tags/ai-pentesting.md>), [aikido-security](<https://devfeed.tech/tags/aikido-security.md>), [android](<https://devfeed.tech/tags/android.md>), [android-apps](<https://devfeed.tech/tags/android-apps.md>), [api](<https://devfeed.tech/tags/api.md>), [apis](<https://devfeed.tech/tags/apis.md>), [app](<https://devfeed.tech/tags/app.md>), [autofix](<https://devfeed.tech/tags/autofix.md>), [backend](<https://devfeed.tech/tags/backend.md>), [company-updates](<https://devfeed.tech/tags/company-updates.md>), [ios](<https://devfeed.tech/tags/ios.md>), [launches](<https://devfeed.tech/tags/launches.md>), [pentesting](<https://devfeed.tech/tags/pentesting.md>), [whitebox](<https://devfeed.tech/tags/whitebox.md>)

### AI overview

Aikido says it now tests iOS alongside Android apps and their shared backend in one whitebox assessment. Its agents install the supplied apps, log in as test users, and attack the apps and API together; using backend source code can sharpen API coverage. Before testing, they produce a recon report and attack plan, and reported findings include risk explanations and reproduction steps. Aikido verifies findings by exploiting them, offers AutoFix pull requests, and supports retests after uploading a rebuilt app. Requirements include connected mobile source code, an APK or IPA, a backend API URL, and test-user credentials. Mobile testing is whitebox only for now; blackbox testing is described as forthcoming. Tests are priced by scope, with cost shown before launch.

### Source excerpt

Aikido now pentests iOS apps. Test your iOS and Android apps and the backend they share in one whitebox AI pentest, with AutoFix and retests. Category: Company Updates

## Ruby2Shell: critical Ruby RCE, affected versions and how to fix it (CVE-2026-132825)

DevFeed: [Ruby2Shell: critical Ruby RCE, affected versions and how to fix it (CVE-2026-132825)](<https://devfeed.tech/articles/ruby2shell-critical-ruby-rce-affected-versions-and-how-to-fix-it-cve-2026-132825-143875.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/ruby2shell>)

Author: Sooraj Shah

Published: 2026-11-19T09:42:00Z

Content type: news

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [cve remediation](<https://devfeed.tech/topics/cve-remediation.md>), [Ruby](<https://devfeed.tech/topics/ruby.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [code-execution](<https://devfeed.tech/tags/code-execution.md>), [critical](<https://devfeed.tech/tags/critical.md>), [cve](<https://devfeed.tech/tags/cve.md>), [discourse](<https://devfeed.tech/tags/discourse.md>), [exploits](<https://devfeed.tech/tags/exploits.md>), [exposed](<https://devfeed.tech/tags/exposed.md>), [github-enterprise-server](<https://devfeed.tech/tags/github-enterprise-server.md>), [gitlab](<https://devfeed.tech/tags/gitlab.md>), [heap-overflow](<https://devfeed.tech/tags/heap-overflow.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [mastodon](<https://devfeed.tech/tags/mastodon.md>), [rce](<https://devfeed.tech/tags/rce.md>), [ruby](<https://devfeed.tech/tags/ruby.md>), [ruby2shell](<https://devfeed.tech/tags/ruby2shell.md>), [threats](<https://devfeed.tech/tags/threats.md>), [upgrade](<https://devfeed.tech/tags/upgrade.md>), [vulnerabilities](<https://devfeed.tech/tags/vulnerabilities.md>), [vulnerabilities-threats](<https://devfeed.tech/tags/vulnerabilities-threats.md>)

### AI overview

Ruby2Shell (CVE-2026-132825) is a critical heap overflow that can enable unauthenticated code execution when applications change the case of untrusted strings in certain ISO-8859-x or Windows-125x encodings; UTF-8 strings are unaffected. The excerpt identifies Ruby 2.4.0-3.3.13, 3.4.0-3.4.12, and 4.0.0-4.0.9 as affected, with Rails and classic-style Sinatra exposed by default. Upgrade to Ruby 3.3.14, 3.4.13, or 4.0.10, restart app servers, and rebuild images or update Ruby version pins. Ruby 3.2 and older have no fix and should move to a supported series. For self-hosted products still bundling vulnerable Ruby, the excerpt recommends restricting server access until vendors release fixes. It reports no known attacks, but notes maintainers know of proof-of-concept exploits.

### Source excerpt

Ruby2Shell is a critical Ruby flaw that can give attackers code execution. See affected versions, whether GitLab is hit, and how to upgrade and fix it. Category: Vulnerabilities & Threats

## PHP critical security release, 19 November (CVE-2026-129289)

DevFeed: [PHP critical security release, 19 November (CVE-2026-129289)](<https://devfeed.tech/articles/php-critical-security-release-19-november-cve-2026-129289-143874.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/php-security-release-19-november-2026-cve-2026-115560>)

Author: Dania Durnas

Published: 2026-11-18T19:28:00Z

Content type: release

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [Security](<https://devfeed.tech/topics/security.md>), [PHP](<https://devfeed.tech/topics/php.md>), [cve remediation](<https://devfeed.tech/topics/cve-remediation.md>), [code security](<https://devfeed.tech/topics/code-security.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [apache](<https://devfeed.tech/tags/apache.md>), [application](<https://devfeed.tech/tags/application.md>), [critical](<https://devfeed.tech/tags/critical.md>), [critical-vulnerabilities](<https://devfeed.tech/tags/critical-vulnerabilities.md>), [cve](<https://devfeed.tech/tags/cve.md>), [cve-2026-115560](<https://devfeed.tech/tags/cve-2026-115560.md>), [cve-2026-129289](<https://devfeed.tech/tags/cve-2026-129289.md>), [cves](<https://devfeed.tech/tags/cves.md>), [cvss](<https://devfeed.tech/tags/cvss.md>), [fixed](<https://devfeed.tech/tags/fixed.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [november-2026](<https://devfeed.tech/tags/november-2026.md>), [patch](<https://devfeed.tech/tags/patch.md>), [php](<https://devfeed.tech/tags/php.md>), [release](<https://devfeed.tech/tags/release.md>), [security](<https://devfeed.tech/tags/security.md>), [threats](<https://devfeed.tech/tags/threats.md>), [vulnerabilities](<https://devfeed.tech/tags/vulnerabilities.md>), [vulnerabilities-threats](<https://devfeed.tech/tags/vulnerabilities-threats.md>)

### AI overview

PHP's 19 November 2026 updates fix 18 CVEs across supported branches: upgrade to 8.5.13, 8.4.28, 8.3.37, or 8.2.36. CVE-2026-129289 (CVSS 9.2) can let an unauthenticated attacker run PHP code through a single request when the front end reuses PHP-FPM connections and an application leaves `php://input` unread; standard nginx and Apache configurations that close connections after each request are not affected by the demonstrated attack. Until upgrading, disable backend connection reuse or close connections after each request, and do not expose PHP-FPM or LSAPI listeners to untrusted networks. Reading the input stream fully is not a reliable mitigation on its own. CVE-2026-115560 requires both a runtime switch of `session.serialize_handler` to `php` before `session_start()` (from `php_binary` or `php_serialize`) and enabled session upload progress. Setting `session.upload_progress.enabled=0` before request parsing and restarting all PHP workers mitigates its published attack path; this does not cover the other critical issues. The November updates also include October's PHP2Shell fixes. The excerpt recommends checking affected versions and CVE advisories, and notes Aikido images back-

### Source excerpt

PHP's 19 November 2026 security release fixes 18 CVEs, including CVE-2026-115560. See fixed versions, whether you're affected and how to patch. Category: Vulnerabilities & Threats

## Aikido's Deep Review finds Java4Shell in the PR that introduced it

DevFeed: [Aikido's Deep Review finds Java4Shell in the PR that introduced it](<https://devfeed.tech/articles/aikido-s-deep-review-finds-java4shell-in-the-pr-that-introduced-it-143872.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/deep-review-finds-java4shell-in-pr>)

Author: Debarshi

Published: 2026-11-18T19:04:00Z

Content type: news

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [ai code review](<https://devfeed.tech/topics/ai-code-review.md>), [Code review](<https://devfeed.tech/topics/code-review.md>), [openjdk](<https://devfeed.tech/topics/openjdk.md>), [software security audit](<https://devfeed.tech/topics/software-security-audit.md>), [Pull Request](<https://devfeed.tech/topics/pull-request.md>), [cve remediation](<https://devfeed.tech/topics/cve-remediation.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>)

Tags: [17](<https://devfeed.tech/tags/17.md>), [2021](<https://devfeed.tech/tags/2021.md>), [2026](<https://devfeed.tech/tags/2026.md>), [accelerated](<https://devfeed.tech/tags/accelerated.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-code](<https://devfeed.tech/tags/ai-code.md>), [analysis](<https://devfeed.tech/tags/analysis.md>), [arm](<https://devfeed.tech/tags/arm.md>), [avx-512-vbmi](<https://devfeed.tech/tags/avx-512-vbmi.md>), [bug](<https://devfeed.tech/tags/bug.md>), [cve](<https://devfeed.tech/tags/cve.md>), [cve-2026-114938](<https://devfeed.tech/tags/cve-2026-114938.md>), [introduced](<https://devfeed.tech/tags/introduced.md>), [java4shell](<https://devfeed.tech/tags/java4shell.md>), [news](<https://devfeed.tech/tags/news.md>), [openjdk](<https://devfeed.tech/tags/openjdk.md>), [pr](<https://devfeed.tech/tags/pr.md>), [pull-request](<https://devfeed.tech/tags/pull-request.md>), [review](<https://devfeed.tech/tags/review.md>), [upstream](<https://devfeed.tech/tags/upstream.md>)

### AI overview

Aikido says Deep Review identified the Java4Shell flaw when run against a reconstruction of the 2021 OpenJDK pull request that introduced it. The excerpt explains that malformed Base64 padding could cause integer underflow in the AVX-512 decoder, producing an oversized write past the destination buffer; the vulnerable path is limited to x86-64 CPUs with AVX-512 VBMI support. Deep Review recommended falling back to the safer decoder when fewer than four input bytes remain, matching the upstream fix for this bug. It also reported two other defects, an oversized vector write and incomplete invalid-character validation, which the excerpt says were later corrected upstream. For mitigation, it recommends upgrading to listed patched Java releases; Aikido also offers patched Temurin 17 and 21 images for JDK and JRE use if upgrading is not yet possible, and describes dashboard and Attack Surface exposure checks.

### Source excerpt

We recreated the 2021 OpenJDK pull request behind Java4Shell (CVE-2026-114938) and ran Deep Review on it. It flagged the bug and suggested the upstream fix. Category: News

## nginx2root: Critical nginx vulnerabilities, including one RCE \[CVE-2026-123007, CVE-2026-123634, CVE-2026-123633, CVE-2026-121810, CVE-2026-113425\]

DevFeed: [nginx2root: Critical nginx vulnerabilities, including one RCE \[CVE-2026-123007, CVE-2026-123634, CVE-2026-123633, CVE-2026-121810, CVE-2026-113425\]](<https://devfeed.tech/articles/nginx2root-critical-nginx-vulnerabilities-including-one-rce-cve-2026-123007-cve-2026-123634-cve-2026-123633-cve-2026-121810-cve-2026-113425-143873.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/nginx2root-critical-nginx-rce-vulnerabilities>)

Author: Oscar Sanchez Jr.

Published: 2026-11-18T17:11:00Z

Content type: tutorial

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [nginx](<https://devfeed.tech/topics/nginx.md>), [cve remediation](<https://devfeed.tech/topics/cve-remediation.md>), [Memory Safety](<https://devfeed.tech/topics/memory-safety.md>), [exploit chaining](<https://devfeed.tech/topics/exploit-chaining.md>), [ingress-nginx](<https://devfeed.tech/topics/ingress-nginx.md>)

Tags: [2015](<https://devfeed.tech/tags/2015.md>), [2018](<https://devfeed.tech/tags/2018.md>), [2026](<https://devfeed.tech/tags/2026.md>), [container](<https://devfeed.tech/tags/container.md>), [critical](<https://devfeed.tech/tags/critical.md>), [cve](<https://devfeed.tech/tags/cve.md>), [cve-2026-113425](<https://devfeed.tech/tags/cve-2026-113425.md>), [cve-2026-121810](<https://devfeed.tech/tags/cve-2026-121810.md>), [cve-2026-123007](<https://devfeed.tech/tags/cve-2026-123007.md>), [cve-2026-123633](<https://devfeed.tech/tags/cve-2026-123633.md>), [cve-2026-123634](<https://devfeed.tech/tags/cve-2026-123634.md>), [http-2](<https://devfeed.tech/tags/http-2.md>), [nginx](<https://devfeed.tech/tags/nginx.md>), [nginx-1-31-9](<https://devfeed.tech/tags/nginx-1-31-9.md>), [nginx2root](<https://devfeed.tech/tags/nginx2root.md>), [rce](<https://devfeed.tech/tags/rce.md>), [threats](<https://devfeed.tech/tags/threats.md>), [update](<https://devfeed.tech/tags/update.md>), [vulnerabilities](<https://devfeed.tech/tags/vulnerabilities.md>), [vulnerabilities-threats](<https://devfeed.tech/tags/vulnerabilities-threats.md>)

### AI overview

The excerpt describes five nginx vulnerabilities and a Linux SCTP kernel flaw that can be chained for root access, including from a Docker container. CVE-2026-123007 affects nginx 1.9.5-1.31.8 when reachable HTTP/2 is enabled. Other flaws target resolver subrequests, HTTP/2 or HTTP/3 connections, and mail proxy authentication. Update to nginx 1.31.9 or 1.30.7; distro packages may backport fixes without changing the reported version, so check security advisories. Disabling HTTP/2 mitigates only CVE-2026-123007 and CVE-2026-123633; other temporary mitigations depend on the flaw, including removing request-time DNS, disabling QUIC, or restricting mail proxy listeners. Archived ingress-nginx bundles its own vulnerable nginx and has no upstream fix. Disabling SCTP via modprobe applies only if SCTP is not already loaded or built into the kernel.

### Source excerpt

nginx2root is five critical nginx RCE vulnerabilities, including CVE-2026-123007 in HTTP/2. Check if you're affected and update to nginx 1.31.9 or 1.30.7. Category: Vulnerabilities & Threats

## Pre-auth RCE flaw in Java 17 puts servers at risk: affected versions and how to fix (CVE-2026-114938)

DevFeed: [Pre-auth RCE flaw in Java 17 puts servers at risk: affected versions and how to fix (CVE-2026-114938)](<https://devfeed.tech/articles/pre-auth-rce-flaw-in-java-17-puts-servers-at-risk-affected-versions-and-how-to-fix-cve-2026-114938-138039.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/java-17-rce-vulnerability-cve-2026-114938>)

Author: Sooraj Shah

Published: 2026-11-17T13:10:00Z

Content type: article

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [openjdk](<https://devfeed.tech/topics/openjdk.md>), [cve remediation](<https://devfeed.tech/topics/cve-remediation.md>), [Java](<https://devfeed.tech/topics/java.md>)

Tags: [cve-2026-114938](<https://devfeed.tech/tags/cve-2026-114938.md>), [java-17](<https://devfeed.tech/tags/java-17.md>), [november-2026](<https://devfeed.tech/tags/november-2026.md>), [remote-code-execution](<https://devfeed.tech/tags/remote-code-execution.md>), [vulnerabilities-threats](<https://devfeed.tech/tags/vulnerabilities-threats.md>)

### AI overview

Oracle's November 17, 2026 update covers 11 Java SE patches, but Oracle has not yet confirmed which patch addresses CVE-2026-114938 or published technical details. Early reports describe an unauthenticated RCE in OpenJDK 17.0.6 and later, rated 9.2 under CVSS 4.0; the reported scope and memory-corruption cause remain unconfirmed. Oracle lists Java SE 8, 11, 17, 21, 25 and 27 as affected by vulnerabilities in the update, not necessarily this specific flaw. Apply the Critical Security Patch Update when available; other OpenJDK vendors and Linux distributors issue builds on their own schedules. No workaround is known. Running Java alone does not establish exploitability: Oracle has not disclosed the vulnerable component or attack conditions. Check all runtimes, including bundled JDKs and container images, and prioritize internet-facing services. As of November 17, the text reports no known exploitation or public proof of concept.

### Source excerpt

Java CVE-2026-114938 is a reported unauthenticated remote code execution flaw in Java 17. See affected versions, Oracle's November 2026 patch and the fix. Category: Vulnerabilities & Threats

## Two malicious agent-framework releases hit PyPI in 33 minutes

DevFeed: [Two malicious agent-framework releases hit PyPI in 33 minutes](<https://devfeed.tech/articles/two-malicious-agent-framework-releases-hit-pypi-in-33-minutes-140871.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/two-malicious-agent-framework-releases-hit-pypi-in-33-minutes>)

Author: Hunter Schwartz

Published: 2026-11-17T12:00:00Z

Content type: news

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [malicious packages](<https://devfeed.tech/topics/malicious-packages.md>), [PyPI](<https://devfeed.tech/topics/pypi.md>), [Microsoft Agent Framework](<https://devfeed.tech/topics/microsoft-agent-framework.md>), [Security Attacks](<https://devfeed.tech/topics/security-attacks.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>)

Tags: [agent-framework](<https://devfeed.tech/tags/agent-framework.md>), [ai](<https://devfeed.tech/tags/ai.md>), [azure](<https://devfeed.tech/tags/azure.md>), [azure-devops](<https://devfeed.tech/tags/azure-devops.md>), [github](<https://devfeed.tech/tags/github.md>), [github-actions](<https://devfeed.tech/tags/github-actions.md>), [malicious](<https://devfeed.tech/tags/malicious.md>), [pypi](<https://devfeed.tech/tags/pypi.md>), [python](<https://devfeed.tech/tags/python.md>), [releases](<https://devfeed.tech/tags/releases.md>), [secrets](<https://devfeed.tech/tags/secrets.md>), [stealer](<https://devfeed.tech/tags/stealer.md>), [threats](<https://devfeed.tech/tags/threats.md>), [vulnerabilities](<https://devfeed.tech/tags/vulnerabilities.md>), [vulnerabilities-threats](<https://devfeed.tech/tags/vulnerabilities-threats.md>)

### AI overview

Two malicious Microsoft Agent Framework PyPI releases, 1.26.1 and 1.26.2, published 33 minutes apart on Nov. 17, 2026, added `.pth` startup hooks that ran a credential stealer whenever Python started. It collected environment variables, credential files, Azure managed identity tokens and CI secrets, then sent data to a webhook.site endpoint. Version 1.26.1 targeted Azure DevOps and Azure-hosted workloads; 1.26.2 moved the payload and added GitHub Actions OIDC token theft, including workflows with `id-token: write`. Both releases were removed from PyPI. The text advises checking for the affected versions and hook files, pinning `agent-framework==1.26.0`, rebuilding affected environments, rotating every secret reachable by affected machines, reviewing Entra ID sign-in logs, and searching proxy and DNS logs for the endpoint.

### Source excerpt

Two malicious agent-framework releases on PyPI ran a credential stealer on every Python start, targeting Azure, Azure DevOps, and GitHub Actions secrets. Category: Vulnerabilities & Threats

## Top 8 Upwind Security alternatives for code and cloud security

DevFeed: [Top 8 Upwind Security alternatives for code and cloud security](<https://devfeed.tech/articles/top-8-upwind-security-alternatives-for-code-and-cloud-security-133371.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/upwind-security-alternatives>)

Author: Sooraj Shah

Published: 2026-11-13T15:39:00Z

Content type: comparison

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [cloud security](<https://devfeed.tech/topics/cloud-security.md>), [code security](<https://devfeed.tech/topics/code-security.md>), [workload protection](<https://devfeed.tech/topics/workload-protection.md>), [Static code analysis](<https://devfeed.tech/topics/static-code-analysis.md>), [Falco](<https://devfeed.tech/topics/falco.md>)

Tags: [alternatives](<https://devfeed.tech/tags/alternatives.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [cloud-security](<https://devfeed.tech/tags/cloud-security.md>), [code](<https://devfeed.tech/tags/code.md>), [compare](<https://devfeed.tech/tags/compare.md>), [runtime](<https://devfeed.tech/tags/runtime.md>), [runtime-security](<https://devfeed.tech/tags/runtime-security.md>), [security](<https://devfeed.tech/tags/security.md>), [top](<https://devfeed.tech/tags/top.md>), [upwind](<https://devfeed.tech/tags/upwind.md>), [upwind-security](<https://devfeed.tech/tags/upwind-security.md>)

### AI overview

Upwind is described as a runtime-first cloud security platform whose newer code and testing capabilities are less mature: the text notes no AutoFix pull requests, IDE checks, authenticated DAST, or AI pentesting. Alternatives differ by focus. Aikido combines cloud posture and reachability with code findings, IDE and pull-request workflows, AutoFix, authenticated DAST, and AI pentesting; its runtime sensors require agents on supported systems. Wiz links cloud findings to source and offers code policies and fix PRs, while Sysdig focuses on workload investigation and configured response. Orca starts with agentless cloud assessment; Cortex Cloud targets production response; Sweet documents runtime containment and temporary hotfixes; Aqua offers local runtime enforcement and disconnected deployments; Microsoft Defender for Cloud fits Microsoft-centric estates, with some capabilities plan-dependent or in preview. The comparison is based on vendor documentation, not hands-on tests or benchmarks.

### Source excerpt

Compare eight Upwind alternatives for code, cloud and runtime security, with a side-by-side table and a shortlist for each buying situation. Category: DevSec Tools & Comparisons

## Top 9 Qualys alternatives for infrastructure and application security

DevFeed: [Top 9 Qualys alternatives for infrastructure and application security](<https://devfeed.tech/articles/top-9-qualys-alternatives-for-infrastructure-and-application-security-133370.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/top-qualys-alternatives>)

Author: Sooraj Shah

Published: 2026-11-13T11:56:00Z

Content type: comparison

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [Cloud & Application Security](<https://devfeed.tech/topics/cloud-application-security.md>), [code security](<https://devfeed.tech/topics/code-security.md>), [orca security](<https://devfeed.tech/topics/orca-security.md>), [snyk](<https://devfeed.tech/topics/snyk.md>), [software composition analysis](<https://devfeed.tech/topics/software-composition-analysis.md>), [software security audit](<https://devfeed.tech/topics/software-security-audit.md>), [Cloud APIs](<https://devfeed.tech/topics/cloud-apis.md>), [Security Information and Event Management (SIEM)](<https://devfeed.tech/topics/siem-security.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [alternatives](<https://devfeed.tech/tags/alternatives.md>), [application](<https://devfeed.tech/tags/application.md>), [application-security](<https://devfeed.tech/tags/application-security.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [code](<https://devfeed.tech/tags/code.md>), [compare](<https://devfeed.tech/tags/compare.md>), [comparisons](<https://devfeed.tech/tags/comparisons.md>), [coverage](<https://devfeed.tech/tags/coverage.md>), [infrastructure](<https://devfeed.tech/tags/infrastructure.md>), [qualys](<https://devfeed.tech/tags/qualys.md>), [rapid7](<https://devfeed.tech/tags/rapid7.md>), [remediation](<https://devfeed.tech/tags/remediation.md>), [security](<https://devfeed.tech/tags/security.md>), [snyk](<https://devfeed.tech/tags/snyk.md>), [tenable](<https://devfeed.tech/tags/tenable.md>), [tools](<https://devfeed.tech/tags/tools.md>), [top](<https://devfeed.tech/tags/top.md>), [wiz](<https://devfeed.tech/tags/wiz.md>)

### AI overview

Qualys remains focused on network and host vulnerability assessment and compliance, while its per-asset, module-based pricing, reported agent CPU spikes, and limited new development for Web Application Scanning prompt teams to consider replacing individual capabilities or running alternatives alongside it. The options differ by scope: Tenable and Rapid7 focus on infrastructure assessment; Aikido, Snyk, Wiz, and Orca connect security findings with code or cloud workflows to varying degrees; Invicti specializes in authenticated web and API testing; Intruder combines attack-surface and vulnerability management; Greenbone OpenVAS Community Edition replaces network and host assessment with a self-managed open-source stack. Tradeoffs include add-on costs and gaps in coverage: Aikido lacks network and infrastructure vulnerability management; Snyk lacks host assessment and cloud posture; Orca lacks network assessment and host patching; Invicti lacks infrastructure and cloud posture; and Greenbone requires staff and infrastructure to operate. The document says it relied on customer feedback and vendor documentation, without hands-on tests or performance benchmarks.

### Source excerpt

Compare the top 9 Qualys alternatives for 2026, from Aikido, Tenable and Rapid7 to Wiz, Orca and Snyk, by cloud, code and remediation coverage. Category: DevSec Tools & Comparisons

## Critical Apache HTTP Server vulnerability can allow remote code execution (CVE-2026-119618)

DevFeed: [Critical Apache HTTP Server vulnerability can allow remote code execution (CVE-2026-119618)](<https://devfeed.tech/articles/critical-apache-http-server-vulnerability-can-allow-remote-code-execution-cve-2026-119618-131403.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/cve-2026-119618-apache-http-server-rce>)

Author: Sooraj Shah

Published: 2026-11-13T08:49:00Z

Content type: article

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [HTTP](<https://devfeed.tech/topics/http.md>), [cve remediation](<https://devfeed.tech/topics/cve-remediation.md>)

Tags: [apache-http-server](<https://devfeed.tech/tags/apache-http-server.md>), [cve-2026-119618](<https://devfeed.tech/tags/cve-2026-119618.md>), [http-2](<https://devfeed.tech/tags/http-2.md>), [remote-code-execution](<https://devfeed.tech/tags/remote-code-execution.md>)

### AI overview

CVE-2026-119618 is a critical race condition in Apache HTTP Server's mod_http2: unsynchronized access in assign_headers() can cause heap corruption and unauthenticated remote code execution. Apache HTTP Server 2.4.55-2.4.70 is affected when HTTP/2 is enabled; standalone mod_h2 builds should be upgraded past 2.0.42. Upgrade Apache to 2.4.71 or later, or apply the operating system vendor's fixed package when available. If patching is not yet possible, disable HTTP/2 by removing h2 and h2c and limiting protocols to HTTP/1.1; this reduces exposure but does not fix the flaw. The excerpt says standard HTTP load balancers generally prevent exposure in the usual setup, no in-the-wild exploitation had been reported at publication, and some distributions had not yet shipped fixes.

### Source excerpt

CVE-2026-119618 is a critical mod_http2 RCE in Apache HTTP Server 2.4.55 to 2.4.70. See who is affected, how to fix it, and how to disable HTTP/2. Category: Vulnerabilities & Threats

## Top AI network and infrastructure pentesting tools in 2026

DevFeed: [Top AI network and infrastructure pentesting tools in 2026](<https://devfeed.tech/articles/top-ai-network-and-infrastructure-pentesting-tools-in-2026-130863.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/top-network-infrastructure-pentesting-tools>)

Author: Sooraj Shah

Published: 2026-11-12T14:01:00Z

Content type: comparison

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [network security](<https://devfeed.tech/topics/network-security.md>), [ai security](<https://devfeed.tech/topics/ai-security.md>), [Artificial Intelligence](<https://devfeed.tech/topics/ai.md>), [Azure](<https://devfeed.tech/topics/azure.md>), [on-prem](<https://devfeed.tech/topics/on-prem.md>), [rancher](<https://devfeed.tech/topics/rancher.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [ai](<https://devfeed.tech/tags/ai.md>), [aikido-security](<https://devfeed.tech/tags/aikido-security.md>), [compare](<https://devfeed.tech/tags/compare.md>), [comparisons](<https://devfeed.tech/tags/comparisons.md>), [devsec-tools-comparisons](<https://devfeed.tech/tags/devsec-tools-comparisons.md>), [external](<https://devfeed.tech/tags/external.md>), [infrastructure](<https://devfeed.tech/tags/infrastructure.md>), [network](<https://devfeed.tech/tags/network.md>), [networks](<https://devfeed.tech/tags/networks.md>), [pentesting](<https://devfeed.tech/tags/pentesting.md>), [pentesting-tools](<https://devfeed.tech/tags/pentesting-tools.md>), [security](<https://devfeed.tech/tags/security.md>), [tools](<https://devfeed.tech/tags/tools.md>)

### AI overview

The right infrastructure pentest depends on where the simulated attacker starts: external tests cover internet-facing systems, while internal assumed-breach tests begin from a foothold and can examine lateral movement and Active Directory. The tools differ in coverage and setup. Aikido Security tests external assets discovered from verified domains, uses agents to pursue attack paths across hosts, and retests for new openings; internal and Active Directory testing is described as coming soon. Pentera and Horizon3 NodeZero support internal testing from a host inside the environment, which can require change requests and firewall rules; NodeZero also runs external tests from its cloud. RidgeBot runs as an appliance or virtual machine inside the network, keeping test data there. Hadrian and Escape focus on external testing and do not cover internal lateral movement. The article cautions that some tools rely on scripted attack paths, and recommends checking what the AI decides, what must be installed, how fixes are retested, and how test actions and data are controlled.

### Source excerpt

Compare six AI network and infrastructure pentesting tools. Aikido Security is our pick for external network pentesting; see what fits internal networks. Category: DevSec Tools & Comparisons

## wfx-loot campaign pushes CI secret stealer to tens of thousands of repositories

DevFeed: [wfx-loot campaign pushes CI secret stealer to tens of thousands of repositories](<https://devfeed.tech/articles/wfx-loot-campaign-pushes-ci-secret-stealer-to-tens-of-thousands-of-repositories-130862.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/compromised-github-accounts-push-ci-secret-stealer-to-tens-of-thousands-of-repositories>)

Author: Oliver Smith

Published: 2026-11-12T01:45:00Z

Content type: news

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [Secret Scanning](<https://devfeed.tech/topics/secret-scanning.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>), [HashiCorp Vault](<https://devfeed.tech/topics/hashicorp-vault.md>)

Tags: [campaign](<https://devfeed.tech/tags/campaign.md>), [ci](<https://devfeed.tech/tags/ci.md>), [repositories](<https://devfeed.tech/tags/repositories.md>), [secret](<https://devfeed.tech/tags/secret.md>), [stealer](<https://devfeed.tech/tags/stealer.md>), [threats](<https://devfeed.tech/tags/threats.md>), [vulnerabilities](<https://devfeed.tech/tags/vulnerabilities.md>), [wfx-loot](<https://devfeed.tech/tags/wfx-loot.md>)

### AI overview

The wfx-loot campaign uses stolen GitHub credentials to probe repositories, inject a roughly 1,500-line shell stealer through one of three CI or install-hook variants, then revert changes and delete workflow runs and logs. The payload collects GitHub, cloud, AI-tool and other runner credentials; can exchange tokens for deeper cloud access and establish persistence through AWS or Azure trust changes; and in this campaign sends data to two HTTP callback servers. At least 1,835 repositories reportedly ran it, though impact may be higher and missing artifacts do not rule out theft. Check for the listed branch names, pull request title, commit messages and affected package versions. If it ran, assume workflow-accessible secrets and cloud roles are compromised. The identified Go module versions contain the payload only in workflow files; the document says Go builds do not run those files, so module consumers may not have been affected.

### Source excerpt

A wave of compromised accounts are pushing a CI secret stealer. See if you're affected, what it steals, and what to rotate. Category: Vulnerabilities & Threats

## Even gem list can run someone else's code

DevFeed: [Even gem list can run someone else's code](<https://devfeed.tech/articles/even-gem-list-can-run-someone-else-s-code-121421.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/gem-list-runs-someone-elses-code>)

Author: Hunter Schwartz

Published: 2026-11-05T20:30:00Z

Content type: tutorial

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [code security](<https://devfeed.tech/topics/code-security.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>)

Tags: [aikido-device-protection](<https://devfeed.tech/tags/aikido-device-protection.md>), [aikido-intel](<https://devfeed.tech/tags/aikido-intel.md>), [code](<https://devfeed.tech/tags/code.md>), [data](<https://devfeed.tech/tags/data.md>), [execution](<https://devfeed.tech/tags/execution.md>), [install](<https://devfeed.tech/tags/install.md>), [ruby](<https://devfeed.tech/tags/ruby.md>), [vulnerabilities-threats](<https://devfeed.tech/tags/vulnerabilities-threats.md>)

### AI overview

RubyGems can run gem code during installation through native extension build files, and installed plugins can run again on every gem command, including `gem list`. The excerpt reports four still-installable gems--xmastree, ancient-city-ruby, hacky, and et_phone_home--that send a username, a macOS screenshot, or a beacon to a remote server. Stable RubyGems 4.0.x lacks opt-outs; RubyGems and Bundler 4.1.0.beta3 add settings that must be enabled to skip extension builds and plugin installation. These settings protect install-only steps, but skipped extensions can make gems unavailable to Bundler or the app, and malicious code can still run when a gem is loaded. The excerpt gives commands to check for the named gems, remove affected dependencies, inspect plugins, enable the opt-outs, and review gem build and plugin files before installation.

### Source excerpt

Ruby gems can run code on install and on every gem command, even gem list. Check for four data-exfiltrating gems and turn off install-time execution. Category: Vulnerabilities & Threats

## Deleting a leaked key from your repo doesn't protect you

DevFeed: [Deleting a leaked key from your repo doesn't protect you](<https://devfeed.tech/articles/deleting-a-leaked-key-from-your-repo-doesn-t-protect-you-121422.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/leaked-api-key-revoke-it-immediately>)

Author: Joe Leon

Published: 2026-11-05T17:07:00Z

Content type: article

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [Secrets Management](<https://devfeed.tech/topics/secrets-management.md>), [Secret Scanning](<https://devfeed.tech/topics/secret-scanning.md>), [Tool](<https://devfeed.tech/topics/tool.md>), [CI/CD](<https://devfeed.tech/topics/cicd.md>), [Pull Request](<https://devfeed.tech/topics/pull-request.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [22](<https://devfeed.tech/tags/22.md>), [3](<https://devfeed.tech/tags/3.md>), [about](<https://devfeed.tech/tags/about.md>), [account](<https://devfeed.tech/tags/account.md>), [actions](<https://devfeed.tech/tags/actions.md>), [added](<https://devfeed.tech/tags/added.md>), [admin](<https://devfeed.tech/tags/admin.md>), [agent](<https://devfeed.tech/tags/agent.md>), [breach](<https://devfeed.tech/tags/breach.md>), [doesn-t](<https://devfeed.tech/tags/doesn-t.md>), [git](<https://devfeed.tech/tags/git.md>), [github](<https://devfeed.tech/tags/github.md>), [history](<https://devfeed.tech/tags/history.md>), [key](<https://devfeed.tech/tags/key.md>), [learn](<https://devfeed.tech/tags/learn.md>), [litellm](<https://devfeed.tech/tags/litellm.md>), [news](<https://devfeed.tech/tags/news.md>), [repo](<https://devfeed.tech/tags/repo.md>), [secrets](<https://devfeed.tech/tags/secrets.md>)

### AI overview

Deleting a leaked credential, rewriting Git history, or deleting the repository does not reliably remove it: scanners can inspect old commits, GitHub may serve rewritten commits by hash, and forks can retain deleted commits. Treat a public secret as compromised and rotate it by revoking it, issuing a replacement, and updating its uses. Then review access logs, branch commits, CI/workflow changes, and pipeline runs for prior misuse; treat the situation as an incident if misuse cannot be ruled out. The excerpt describes Betterleaks scanning Git history, Aikido repository scanning with optional history checks, active-key detection and triage, IDE/pre-commit warnings and CI blocking, plus device scanning that keeps file contents local.

### Source excerpt

A leaked admin key on GitHub, deleted within hours, still led to a breach months later at LiteLLM. Learn why Git history keeps secrets and why you have to rotate. Category: News

## Surviving cybersecurity until the end of 2026

DevFeed: [Surviving cybersecurity until the end of 2026](<https://devfeed.tech/articles/surviving-cybersecurity-until-the-end-of-2026-119297.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/cybersecurity-surviving-vulnerabilities-malware-best-practices>)

Author: Chris Holman

Published: 2026-11-04T21:06:00Z

Content type: article

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [Cybersecurity](<https://devfeed.tech/topics/cybersecurity.md>), [container-security](<https://devfeed.tech/topics/container-security.md>)

Tags: [2019](<https://devfeed.tech/tags/2019.md>), [2026](<https://devfeed.tech/tags/2026.md>), [2fa](<https://devfeed.tech/tags/2fa.md>), [adversary](<https://devfeed.tech/tags/adversary.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-code](<https://devfeed.tech/tags/ai-code.md>), [ai-tools](<https://devfeed.tech/tags/ai-tools.md>), [api](<https://devfeed.tech/tags/api.md>), [api-keys](<https://devfeed.tech/tags/api-keys.md>), [asm](<https://devfeed.tech/tags/asm.md>), [attack](<https://devfeed.tech/tags/attack.md>), [attack-surface](<https://devfeed.tech/tags/attack-surface.md>), [binance](<https://devfeed.tech/tags/binance.md>), [compromise](<https://devfeed.tech/tags/compromise.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [emudhra](<https://devfeed.tech/tags/emudhra.md>), [litellm](<https://devfeed.tech/tags/litellm.md>), [news](<https://devfeed.tech/tags/news.md>), [php2shell](<https://devfeed.tech/tags/php2shell.md>), [practices](<https://devfeed.tech/tags/practices.md>)

### AI overview

The article recommends assuming a developer laptop, vendor, package, or CI runner will eventually be compromised, then limiting how far an attacker can move. Its six practices are to use short-lived, narrowly scoped credentials and audit tokens; reduce inherited code with minimal or hardened images, dependency tracking, and prompt patching; inventory and remove unnecessary vendors and integrations; find and take unneeded services offline; control software and extensions on user machines while keeping secrets off disk; and routinely test exposure through attack surface management, pentests, and breach exercises. A staged checklist calls for a 48-hour package-age minimum and internet exposure review this week, then credential rotation, inventories, auto-shutdown, and critical patch targets this month, followed by broader testing and reviews before 2027.

### Source excerpt

After Binance, LiteLLM, PHP2Shell and eMudhra, here are six practices and a checklist to limit what the next compromise can reach before 2027. Category: News

## Critical Linux kernel TCP vulnerability can allow remote code execution (CVE-2026-98432)

DevFeed: [Critical Linux kernel TCP vulnerability can allow remote code execution (CVE-2026-98432)](<https://devfeed.tech/articles/critical-linux-kernel-tcp-vulnerability-can-allow-remote-code-execution-cve-2026-98432-112422.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/linux-kernel-tcp-rce-vulnerability-cve-2026-98432>)

Author: Felix Garriau

Published: 2026-11-02T09:00:00Z

Content type: tutorial

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [Linux Kernel](<https://devfeed.tech/topics/linux-kernel.md>), [Kernel](<https://devfeed.tech/topics/kernel.md>), [Linux](<https://devfeed.tech/topics/linux.md>)

Tags: [cve-2026-98432](<https://devfeed.tech/tags/cve-2026-98432.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [linux-kernel](<https://devfeed.tech/tags/linux-kernel.md>), [patch](<https://devfeed.tech/tags/patch.md>), [remote-code-execution](<https://devfeed.tech/tags/remote-code-execution.md>), [tcp](<https://devfeed.tech/tags/tcp.md>), [vulnerabilities-threats](<https://devfeed.tech/tags/vulnerabilities-threats.md>), [vulnerability](<https://devfeed.tech/tags/vulnerability.md>)

### AI overview

CVE-2026-98432 is a use-after-free race in Linux IPv4 TCP connection setup, present since kernel 4.4. The kernel CVE team rates it 9.8; Anthropic reports a demonstrated remote root exploit on a server with an MPTCP listener, while AWS describes remote code execution under specific conditions. The mechanism involves IP options freed while another CPU may still read them; the upstream fix uses kfree_rcu(). Patch the kernel and reboot, then verify the running version with uname -r against the distribution advisory, which may include backports. Check every VM and container node. If patching is not yet possible, disabling MPTCP, dropping IPv4 packets with IP options, or terminating TCP at a separate load balancer can reduce exposure but does not fix the bug. Exposure and patch availability vary by platform and vendor; the text reports no vendor advisories of in-the-wild attacks as of October 29 and notes unresolved questions about exploit reproducibility and whether MPTCP is required.

### Source excerpt

CVE-2026-98432 is a critical Linux kernel TCP flaw rated 9.8 that can allow remote code execution. See who is affected, how to patch and what to do now. Category: Vulnerabilities & Threats

## AI agent security in Kubernetes: Why prompt injection isn't the right term

DevFeed: [AI agent security in Kubernetes: Why prompt injection isn't the right term](<https://devfeed.tech/articles/ai-agent-security-in-kubernetes-why-prompt-injection-isn-t-the-right-term-109778.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/ai-agent-security-in-kubernetes-prompt-injection>)

Author: Nicholas Thomson

Published: 2026-10-30T19:31:00Z

Content type: article

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [LLM security](<https://devfeed.tech/topics/llm-security.md>), [ai security](<https://devfeed.tech/topics/ai-security.md>), [prompt injection](<https://devfeed.tech/topics/prompt-injection.md>), [Security](<https://devfeed.tech/topics/security.md>), [human in the loop agent](<https://devfeed.tech/topics/human-in-the-loop-agent.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agent-security](<https://devfeed.tech/tags/agent-security.md>), [agentic-ai-security](<https://devfeed.tech/tags/agentic-ai-security.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [ai-agent-security](<https://devfeed.tech/tags/ai-agent-security.md>), [attack-path-analysis](<https://devfeed.tech/tags/attack-path-analysis.md>), [best](<https://devfeed.tech/tags/best.md>), [best-practices](<https://devfeed.tech/tags/best-practices.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [developer](<https://devfeed.tech/tags/developer.md>), [developer-advocate](<https://devfeed.tech/tags/developer-advocate.md>), [engineer](<https://devfeed.tech/tags/engineer.md>), [google](<https://devfeed.tech/tags/google.md>), [guides](<https://devfeed.tech/tags/guides.md>), [isn-t](<https://devfeed.tech/tags/isn-t.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [kubernetes-security](<https://devfeed.tech/tags/kubernetes-security.md>), [llm-security](<https://devfeed.tech/tags/llm-security.md>), [prompt](<https://devfeed.tech/tags/prompt.md>), [prompt-injection](<https://devfeed.tech/tags/prompt-injection.md>), [prompt-poisoning](<https://devfeed.tech/tags/prompt-poisoning.md>), [rbac](<https://devfeed.tech/tags/rbac.md>), [right](<https://devfeed.tech/tags/right.md>), [secrets-detection](<https://devfeed.tech/tags/secrets-detection.md>), [security](<https://devfeed.tech/tags/security.md>), [technical-educational](<https://devfeed.tech/tags/technical-educational.md>)

### AI overview

Kubernetes agents can be hijacked through attacker-controlled text in inputs such as application logs, which the model may treat like instructions because it processes them alongside system prompts and conversation history. Broad read access then determines what a compromised agent can expose: read-only mode can prevent destructive cluster actions, but does not prevent it from reading secrets, environment variables, or poisoned logs. The excerpt recommends limiting RBAC and cloud identity permissions to the agent's needs and having a human verify its actions, while acknowledging that reliable trust and verification remain unresolved. It describes checks including testing whether poisoned retrieved documents or tool output can override instructions or trigger unauthorized tool calls; reviewing infrastructure-as-code permissions and runtime RBAC and workload identity; scanning secrets and container images; and tracing attack paths to sensitive systems. The term "prompt poisoning" is preferred because the excerpt says LLMs lack a SQL-style mechanism for separating instructions from input.

### Source excerpt

Abdel Sghiouar, Google engineer and senior cloud developer advocate, thinks prompt injection should be called prompt poisoning Category: Guides & Best Practices

## Binance hack: how a compromised developer laptop can lead to a $6.3 billion breach

DevFeed: [Binance hack: how a compromised developer laptop can lead to a $6.3 billion breach](<https://devfeed.tech/articles/binance-hack-how-a-compromised-developer-laptop-can-lead-to-a-6-3-billion-breach-109148.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/binance-hack-developer-laptop>)

Author: Sooraj Shah

Published: 2026-10-30T13:13:00Z

Content type: news

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [Binance](<https://devfeed.tech/topics/binance.md>), [Credential theft](<https://devfeed.tech/topics/credential-theft.md>), [Post Mortem](<https://devfeed.tech/topics/post-mortem.md>)

Tags: [attack](<https://devfeed.tech/tags/attack.md>), [binance](<https://devfeed.tech/tags/binance.md>), [binance-hack](<https://devfeed.tech/tags/binance-hack.md>), [device-protection](<https://devfeed.tech/tags/device-protection.md>), [eip-7702](<https://devfeed.tech/tags/eip-7702.md>), [funds](<https://devfeed.tech/tags/funds.md>), [laptop](<https://devfeed.tech/tags/laptop.md>)

### AI overview

Binance says its investigation initially pointed to a compromised DevOps laptop, then found additional breach points it has not described. It reports about $6.3 billion in impact, with roughly $3 billion frozen, and says it will cover losses. Co-CEO Richard Teng said the attacker used Ethereum's EIP-7702 to gain approval to act on a hot wallet outside Binance's normal controls; Binance has reported no evidence of a private-key leak and has not explained how the laptop was compromised. The article outlines a possible route from malware on a developer device through stored credentials or session tokens to internal systems and a withdrawal service, while noting other systems could also have been breached. It recommends short-lived credentials, independent approval for high-value actions, alerts on unusual credential use, and managed controls for contractor and personal devices with production access. Aikido's device agent is described as blocking malicious packages, extensions, and AI tools, with a default 48-hour holdback for newly published items. TRM Labs and Chainalysis attribute the incident to TraderTraitor; Binance has not confirmed that attribution.

### Source excerpt

The Binance hack may have started with a compromised developer laptop, Binance says. See how that can lead to stolen funds and how to protect against it. Category: News

## What are shell vulnerabilities? PHP2Shell, Node2Shell (HTTP2Shell), React2Shell and Log4Shell explained

DevFeed: [What are shell vulnerabilities? PHP2Shell, Node2Shell (HTTP2Shell), React2Shell and Log4Shell explained](<https://devfeed.tech/articles/what-are-shell-vulnerabilities-php2shell-node2shell-http2shell-react2shell-and-log4shell-explained-107251.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/what-are-shell-vulnerabilities>)

Author: Nicholas Thomson

Published: 2026-10-29T19:07:00Z

Content type: article

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [High Profile Threats](<https://devfeed.tech/topics/high-profile-threats.md>), [code security](<https://devfeed.tech/topics/code-security.md>), [known exploitable vulnerabilities](<https://devfeed.tech/topics/known-exploitable-vulnerabilities.md>), [Security research](<https://devfeed.tech/topics/security-research.md>), [software composition analysis](<https://devfeed.tech/topics/software-composition-analysis.md>), [container images](<https://devfeed.tech/topics/container-images.md>), [cve remediation](<https://devfeed.tech/topics/cve-remediation.md>)

Tags: [2shell-vulnerabilities](<https://devfeed.tech/tags/2shell-vulnerabilities.md>), [ai-vulnerability-discovery](<https://devfeed.tech/tags/ai-vulnerability-discovery.md>), [black-october](<https://devfeed.tech/tags/black-october.md>), [broccoli2shell](<https://devfeed.tech/tags/broccoli2shell.md>), [brotli-vulnerability](<https://devfeed.tech/tags/brotli-vulnerability.md>), [coordinated-disclosure](<https://devfeed.tech/tags/coordinated-disclosure.md>), [cve-2021-44228](<https://devfeed.tech/tags/cve-2021-44228.md>), [cve-2025-55182](<https://devfeed.tech/tags/cve-2025-55182.md>), [cve-2026-109618](<https://devfeed.tech/tags/cve-2026-109618.md>), [cve-2026-109619](<https://devfeed.tech/tags/cve-2026-109619.md>), [cve-2026-111497](<https://devfeed.tech/tags/cve-2026-111497.md>), [http2shell](<https://devfeed.tech/tags/http2shell.md>), [insecure-deserialization](<https://devfeed.tech/tags/insecure-deserialization.md>), [jndi-lookup-injection](<https://devfeed.tech/tags/jndi-lookup-injection.md>), [log4shell](<https://devfeed.tech/tags/log4shell.md>), [node-js-vulnerability](<https://devfeed.tech/tags/node-js-vulnerability.md>), [node2shell](<https://devfeed.tech/tags/node2shell.md>), [out-of-bounds-write](<https://devfeed.tech/tags/out-of-bounds-write.md>), [php-vulnerability](<https://devfeed.tech/tags/php-vulnerability.md>), [php2shell](<https://devfeed.tech/tags/php2shell.md>), [react2shell](<https://devfeed.tech/tags/react2shell.md>), [remote-code-execution](<https://devfeed.tech/tags/remote-code-execution.md>), [shell](<https://devfeed.tech/tags/shell.md>), [shell-vulnerability](<https://devfeed.tech/tags/shell-vulnerability.md>), [unauthenticated-rce](<https://devfeed.tech/tags/unauthenticated-rce.md>), [use-after-free](<https://devfeed.tech/tags/use-after-free.md>), [vulnerabilities](<https://devfeed.tech/tags/vulnerabilities.md>), [vulnerabilities-threats](<https://devfeed.tech/tags/vulnerabilities-threats.md>)

### AI overview

Shell vulnerabilities are remote code execution flaws caused when attacker-controlled input reaches unsafe code in underlying libraries, frameworks, runtimes or decoders; the text says these can affect teams even when their own application code is careful, and often require no login. It distinguishes memory corruption bugs PHP2Shell and Node2Shell from logic flaws Log4Shell and React2Shell. PHP2Shell involves use-after-free conditions in PHP request handling; Node2Shell is a Brotli buffer overrun that can affect servers and clients, with code execution demonstrated but inconsistent and crashes or leaks described as likelier. Log4Shell abused Log4j lookups in logged user text, while React2Shell exploited Flight deserialization in React Server Components; both had follow-up fixes after initial patches. The text recommends identifying vulnerable dependencies and container images, upgrading to fixed releases, and tracking ecosystem-specific Brotli patches, which arrive on different schedules. It also describes Aikido scanning, patched images for some versions lacking upstream fixes, and a public-URL checker for PHP2Shell.

### Source excerpt

We explain what shell vulnerabilities are and how to manage them: PHP2Shell, Node2Shell (HTTP2Shell), React2Shell and Log4Shell Category: Guides & Best Practices

## The day after tomorrow: After PHP2Shell, it gets worse, then defenders win

DevFeed: [The day after tomorrow: After PHP2Shell, it gets worse, then defenders win](<https://devfeed.tech/articles/the-day-after-tomorrow-after-php2shell-it-gets-worse-then-defenders-win-106472.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/php2shell-day-after-tomorrow>)

Author: Sooraj Shah

Published: 2026-10-29T13:00:00Z

Content type: news

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [code security](<https://devfeed.tech/topics/code-security.md>), [cve remediation](<https://devfeed.tech/topics/cve-remediation.md>), [known exploitable vulnerabilities](<https://devfeed.tech/topics/known-exploitable-vulnerabilities.md>), [Frontier AI](<https://devfeed.tech/topics/frontier-ai.md>), [container images](<https://devfeed.tech/topics/container-images.md>), [human review](<https://devfeed.tech/topics/human-review.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [brotli](<https://devfeed.tech/tags/brotli.md>), [code-execution](<https://devfeed.tech/tags/code-execution.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [defender](<https://devfeed.tech/tags/defender.md>), [php2shell](<https://devfeed.tech/tags/php2shell.md>), [vulnerabilities](<https://devfeed.tech/tags/vulnerabilities.md>)

### AI overview

The excerpt argues that AI can make overlooked bugs in foundational software easier to turn into exploits, as illustrated by PHP2Shell and a Brotli flaw; it reports no mass compromise at the time described. It recommends combining strong security architecture, which limits damage, with rapid remediation. Practical steps include inventorying software dependencies--even those in container and CI images or vendor products--checking code and testing applications and external networks before release and after changes, and routing findings to fixes a person can review. If no upgrade is available or one would break things, it suggests applying the fix to the version already in use. The excerpt says supported PHP versions had fixes and some sites went offline, while warning that the near term could remain difficult.

### Source excerpt

PHP2Shell and Brocolli2Shell may be harbingers of many vulnerabilities to come. But we can be ready Category: News

## Ghost in the PHP2Shell: Why coordinated disclosure and patching will start moving faster

DevFeed: [Ghost in the PHP2Shell: Why coordinated disclosure and patching will start moving faster](<https://devfeed.tech/articles/ghost-in-the-php2shell-why-coordinated-disclosure-and-patching-will-start-moving-faster-105779.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/ghost-php2shell-coordinated-disclosure-faster>)

Author: Dania Durnas

Published: 2026-10-29T04:44:47.605032Z

Content type: opinion

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [incident](<https://devfeed.tech/topics/incident.md>), [container image compliance](<https://devfeed.tech/topics/container-image-compliance.md>), [container images](<https://devfeed.tech/topics/container-images.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [disclosure](<https://devfeed.tech/tags/disclosure.md>), [exploit](<https://devfeed.tech/tags/exploit.md>), [patching](<https://devfeed.tech/tags/patching.md>), [php2shell](<https://devfeed.tech/tags/php2shell.md>)

### AI overview

PHP2Shell illustrates how AI-assisted vulnerability discovery and exploit development can compress response timelines: Calif and Wiz independently found the flaws, Aikido's AI Code Analysis rediscovered one within hours, and Trail of Bits built an exploit with Codex in a day. Leaked CVE identifiers prompted PHP to move its coordinated patch release forward, but the article reports exploitation attempts before the public pre-announcement and widespread attempts afterward. Shorter, tighter embargoes may reduce leak exposure, while giving downstream distributors, hosting providers, and protective services less time to prepare; Cloudflare, with advance notice, had WAF rules ready before the patch. As of October 27, Censys counted about 11.2 million PHP services on fixed versions and 32.5 million still affected among those reporting versions, with over 60 million not reporting a version. The article advises teams to inventory runtimes and dependencies, use timely incident sources that provide mitigations and deployable fixes, prepare for rapid patching, and lock dependencies rather than automatically upgrading everything. It describes Aikido's backported fixed images as one way to patch

### Source excerpt

PHP2Shell showed how fast AI turns a bug into an exploit. Here's why the disclosure should never happen this way again, and why it will anyway. Category: News

## Aikido launches external network pentesting

DevFeed: [Aikido launches external network pentesting](<https://devfeed.tech/articles/aikido-launches-external-network-pentesting-105778.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/aikido-launches-external-network-pentesting>)

Author: Mario Popescu

Published: 2026-10-28T10:00:00Z

Content type: release

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [network security](<https://devfeed.tech/topics/network-security.md>), [AI Networking](<https://devfeed.tech/topics/ai-networking.md>), [Virtual Private Network](<https://devfeed.tech/topics/vpn.md>), [ssh](<https://devfeed.tech/topics/ssh.md>), [Cloudflare](<https://devfeed.tech/topics/cloudflare.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [2018](<https://devfeed.tech/tags/2018.md>), [2026](<https://devfeed.tech/tags/2026.md>), [3](<https://devfeed.tech/tags/3.md>), [a16z](<https://devfeed.tech/tags/a16z.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [attack](<https://devfeed.tech/tags/attack.md>), [company](<https://devfeed.tech/tags/company.md>), [company-updates](<https://devfeed.tech/tags/company-updates.md>), [exploit](<https://devfeed.tech/tags/exploit.md>), [external](<https://devfeed.tech/tags/external.md>), [internet](<https://devfeed.tech/tags/internet.md>), [launches](<https://devfeed.tech/tags/launches.md>), [network](<https://devfeed.tech/tags/network.md>), [pentesting](<https://devfeed.tech/tags/pentesting.md>), [report](<https://devfeed.tech/tags/report.md>), [systems](<https://devfeed.tech/tags/systems.md>), [updates](<https://devfeed.tech/tags/updates.md>)

### AI overview

Aikido describes an external network pentest that starts with a domain, maps internet-facing hosts, and uses adaptive AI agents to test each host for known vulnerabilities, misconfigurations, and exposed secrets. The scope includes publicly reachable infrastructure and unauthenticated app surfaces; it excludes logged-in app logic and internal systems. Agents must support findings with evidence validated and reproduced by other agents, and reports can show chained access paths. Customers verify ownership, receive a quote, then launch; findings include fix suggestions, with retests available for up to six months. Continuous testing checks newly live hosts and is priced yearly based on network size at the first pentest. The article says safeguards include reviewing actions for disruption, stopping after proven exploitation pending human confirmation, not storing downloaded data, checking target ownership, and an instant stop control.

### Source excerpt

Aikido launches external network pentesting: AI agents attack your internet-facing systems from the outside and report only findings they actually exploit. Category: Company Updates

## How to patch PHP2Shell with free Aikido Images

DevFeed: [How to patch PHP2Shell with free Aikido Images](<https://devfeed.tech/articles/how-to-patch-php2shell-with-free-aikido-images-101316.md>)

Original publisher: [Read original article](<https://www.aikido.dev/blog/how-to-patch-php2shell-with-free-aikido-images>)

Author: Madeline Lawrence

Published: 2026-10-26T22:15:00Z

Content type: tutorial

Language: en

Sources: [Aikido Security's Blog](<https://devfeed.tech/sources/aikido-security-s-blog.md>)

Topics: [chainguard cve remediation](<https://devfeed.tech/topics/chainguard-cve-remediation.md>), [container image compliance](<https://devfeed.tech/topics/container-image-compliance.md>)

Tags: [2026](<https://devfeed.tech/tags/2026.md>), [7-4](<https://devfeed.tech/tags/7-4.md>), [cloudflare](<https://devfeed.tech/tags/cloudflare.md>), [code-execution](<https://devfeed.tech/tags/code-execution.md>), [container](<https://devfeed.tech/tags/container.md>), [container-images](<https://devfeed.tech/tags/container-images.md>), [container-registry](<https://devfeed.tech/tags/container-registry.md>), [containers](<https://devfeed.tech/tags/containers.md>), [critical](<https://devfeed.tech/tags/critical.md>), [cve](<https://devfeed.tech/tags/cve.md>), [cve-2026-109618](<https://devfeed.tech/tags/cve-2026-109618.md>), [cve-2026-109619](<https://devfeed.tech/tags/cve-2026-109619.md>), [free](<https://devfeed.tech/tags/free.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [images](<https://devfeed.tech/tags/images.md>), [patch](<https://devfeed.tech/tags/patch.md>), [php](<https://devfeed.tech/tags/php.md>), [php2shell](<https://devfeed.tech/tags/php2shell.md>), [product-updates](<https://devfeed.tech/tags/product-updates.md>), [updates](<https://devfeed.tech/tags/updates.md>)

### AI overview

PHP2Shell comprises two critical unauthenticated remote code execution flaws (CVE-2026-109618 and CVE-2026-109619); the article says exploitation has begun. PHP's fixes cover versions 8.2-8.5, while Aikido offers free container images with fixes backported across release lines 5.5-8.5 except 8.0, allowing many users to stay on their current line. The text gives Docker pull and Dockerfile examples and says to rebuild and redeploy every PHP image, restart workers, keep session.upload_progress.enabled=0, check the public site with Aikido's checker, and investigate possible compromise. PHP 8.0 and some variants are unavailable; users whose version is not listed are advised to update or seek a vendor backport. PHP's security team also provides source backports for 7.4-8.1.

### Source excerpt

Patch PHP2Shell (CVE-2026-109618 and CVE-2026-109619) with free Aikido Images for PHP 5.5 to 8.5. Stay on your current version with backported fixes. Category: Product Updates

[Next page](<https://devfeed.tech/sources/aikido-security-s-blog.md?cursor=WyIyMDI2LTEwLTI2VDIyOjE1OjAwKzAwOjAwIiwgIjYwNmMwZTI0LWU3MzYtNDNhMS04N2FkLWVjOTQ3YjdkMDZmMiJd>)