https://threatfrontier.com/articles/underdog-saluki-27b-qwen3-8-27b-2-bit-gguf-16gb-gpu ThreatFrontier.com en 2026-10-13T13:43:53.819Z Underdog Saluki 27B Squeezes Qwen3.8-27B Into a 7.89 GB 2-Bit GGUF for 16 GB GPUs https://threatfrontier.com/articles/aleph-alpha-kolibri-1-78b-moe-open-weights-single-h200 ThreatFrontier.com en 2026-10-13T13:40:37.418Z Aleph Alpha's Kolibri-1 Puts a 78B Open-Weight MoE With 1M-Token Context on a Single H200 https://threatfrontier.com/articles/phishing-emails-hidden-prompt-injection-ai-email-assistants-barracuda ThreatFrontier.com en 2026-10-13T13:39:50.206Z Phishing Emails Now Carry Hidden Instructions for AI Assistants as Well as Lures for People https://threatfrontier.com/articles/gsa-llm-clause-552-239-7001-7-day-model-change-notice-no-training ThreatFrontier.com en 2026-10-13T13:39:17.038Z GSA's New AI Contract Clause: 7-Day Warning on Model Regressions and No Training on Agency Data https://threatfrontier.com/articles/gemini-api-deprecations-3-7-flash-deep-research-october-23-nano-banana-2-1 ThreatFrontier.com en 2026-10-13T13:36:02.385Z Gemini API Migration Checklist: 3.7 and 3.5 Flash Deprecated, Deep Research Preview Ends Oct 23 https://threatfrontier.com/articles/cyberxero-claude-code-51-agents-pentagi-ukraine-utilities-wordpress ThreatFrontier.com en 2026-10-13T13:35:55.462Z CyberXero Broker Ran 51 Claude Code Agents Against Ukrainian Utilities and WordPress Sites https://threatfrontier.com/articles/tofu-mcp-server-fuzzing-253-bugs-118-servers ThreatFrontier.com en 2026-10-13T13:35:14.420Z Fuzzing 273 MCP Servers Found 253 Verified Bugs, and Most Tools Trust Whatever the Model Sends https://threatfrontier.com/articles/vllm-0-31-upgrade-notes-preload-criu-snapshots-cuda-13-breaking-changes ThreatFrontier.com en 2026-10-13T13:35:10.017Z vLLM v0.31.0 Upgrade Notes: Preload Daemon, CRIU Snapshots, CUDA 13 Default and the Breaking Changes https://threatfrontier.com/articles/ollama-0-40-mlx-default-background-model-upgrade-disk-backups ThreatFrontier.com en 2026-10-13T13:35:05.782Z Ollama 0.40 Makes MLX the Default on Macs, and 0.40.2 Keeps a Backup of Every Upgraded Model https://threatfrontier.com/articles/gpt-5-5-retires-chatgpt-codex-october-14-replacement-by-plan ThreatFrontier.com en 2026-10-13T13:35:01.462Z GPT-5.5 Retires From ChatGPT and Codex on October 14: What Replaces It on Each Plan https://threatfrontier.com/articles/chatgpt-floor-plan-to-scale-we-tested-claude-and-gemini ThreatFrontier.com en 2026-10-13T13:00:00.000Z ChatGPT Floor Plan to Scale? We Tested Claude and Gemini https://threatfrontier.com/articles/sap-cloud-sdk-python-cve-2026-89111-oauth2-ssrf-october-patch-day ThreatFrontier.com en 2026-10-13T10:43:02.182Z SAP Cloud SDK for Python OAuth2 SSRF (CVE-2026-89111, CVSS 10.0) Leads October Patch Day https://threatfrontier.com/articles/ollama-pull-api-ssrf-cve-2026-108836-exposed-servers-public-exploit ThreatFrontier.com en 2026-10-13T10:32:53.429Z Ollama Pull API SSRF (CVE-2026-108836): Medium-Severity Flaw in Versions Up to 0.34.1, With a Public Exploit https://threatfrontier.com/articles/meta-muse-wonder-api-price-vs-opus-5-5-wonder-full-muse-code-plans ThreatFrontier.com en 2026-10-13T02:44:16.438Z Meta's Muse Wonder Costs the Same as Opus 5.5 on the API; Wonder Full Is Plan-Only https://threatfrontier.com/articles/bricksforge-cve-2026-85097-unauthenticated-upload-rce-exploited ThreatFrontier.com en 2026-10-13T01:33:59.359Z Bricksforge CVE-2026-85097: Unauthenticated PHP Upload Under Attack Since 7 October https://threatfrontier.com/articles/sonicwall-sma1000-cve-2026-102255-pre-auth-ssrf-exploit-attempts ThreatFrontier.com en 2026-10-13T01:26:43.496Z SonicWall SMA1000 CVSS 10 Pre-Auth SSRF CVE-2026-102255 Draws Exploit Attempts Within Two Days https://threatfrontier.com/articles/floorgen-review-free-floor-plan-maker-no-sign-up ThreatFrontier.com en 2026-10-12T13:00:00.000Z FloorGen Review: A Free Floor Plan Maker With No Sign-Up and an MCP Server https://threatfrontier.com/articles/fake-claude-installer-google-bing-ad-swaps-copied-command-zsh ThreatFrontier.com en 2026-10-12T02:45:54.053Z Google ad that points to Bing leads to a fake Claude installer that swaps the command you copy https://threatfrontier.com/articles/ahsaycbs-cve-2026-105133-cve-2026-105134-exploited-unpatched ThreatFrontier.com en 2026-10-12T01:26:55.138Z Attackers chain two unpatched AhsayCBS flaws to get SYSTEM webshells and a hidden Monero miner https://threatfrontier.com/articles/cisa-kev-struts-proftpd-bind-strapi-aa26-281a ThreatFrontier.com en 2026-10-11T20:45:40.597Z CISA adds four old Struts, ProFTPD, BIND and Strapi flaws to KEV, tied to China-linked intrusions https://threatfrontier.com/articles/artex-claude-code-open-directories-south-korea-finance-crowdstrike ThreatFrontier.com en 2026-10-11T20:33:17.657Z Open directories exposed an attacker's Claude Code sessions: how ARTEX was used against Korean financial firms https://threatfrontier.com/articles/github-copilot-cli-encrypted-injection-not-a-vulnerability ThreatFrontier.com en 2026-10-11T20:33:11.644Z GitHub Says Encrypted Copilot CLI Secret Theft Is Not a Vulnerability; Adversa Disagrees https://threatfrontier.com/articles/anthropic-claude-unintended-actions-real-websites ThreatFrontier.com en 2026-10-11T20:27:20.684Z Anthropic Says Claude Took Unintended Actions on Real Websites, Including US Government Ones