Threat intelligence editor

Alex Kim

Articles: 235
Desks: 7

Focuses on cloud identity, incident response, and exploitation trends.

CloudIncident responseIdentity

Coverage by Alex Kim

MindSearch CVE-2026-105135 has a public PoC and no patch; CVSS v3.1 10.0, unauthenticated. Path runs from user query to planner agent to code execution.

highAI Security Cve 2026 105135

MindSearch code-injection flaw (CVE-2026-105135) has a public PoC and no fix

CVE-2026-105135, a code-injection flaw in InternLM MindSearch's Planner Agent, has a public PoC and no known fix. CVSS v3.1 10.0 (VulDB).

Oct 4 · 6 min

CVSS 8.7 poster: Citrix NetScaler SAML flaw CVE-2026-88779 on CISA KEV

highExploits Cve 2026 88779

Citrix NetScaler SAML Memory Flaw CVE-2026-88779 Added to CISA KEV, and Last Month's Patch Does Not Cover It

CVE-2026-88779, a SAML-only memory overflow in Citrix NetScaler, is on CISA KEV with a 7 October deadline. September's patch doesn't cover it.

Oct 4 · 5 min

MCP fetch server SSRF CVE-2026-104120: no fix, CVSS 3.1 7.3 scored by VulDB, fix PR #4890 still a draft

elevatedAI Security Cve 2026 104120

MCP Fetch Server SSRF (CVE-2026-104120) Has a Public Exploit and No Fix

CVE-2026-104120 lets an agent steer the official MCP fetch tool to internal and cloud metadata addresses. A public exploit exists; fix PR #4890 is unmerged.

Oct 3 · 6 min

Data graphic: n8n queue-mode flaw CVE-2026-103251, where write access to Redis can install any npm package on every instance, shown with a CVSS v4.0 High score of 7.5. Redis write access is the precondition; no known exploitation; not in CISA KEV; fixed in 2.40.1.

elevatedAI Security Cve 2026 103251

n8n Queue Mode: Redis Write Access Can Install Any npm Package on Every Instance (CVE-2026-103251)

n8n's queue-mode handler skipped name, permission, checksum and npm safety checks, so Redis write access could install any npm package cluster-wide. Fixed; no known exploitation.

Oct 3 · 5 min

Data graphic: a dot plot of selected open-model coding plans by monthly price in October 2026. OpenCode Go, Chutes Plus and Command Code GOAT which adds a processing fee sell a flat $10 plan, while the others run from $18 Z.ai GLM Lite to $50 Alibaba Qwen Pro , and MiniMax Plus at $22 is closed.

watchResearch Opencode Go

The $10 Open-Model Coding Plan in October 2026: Three Real Options, Six Near Misses, and the Math

Only OpenCode Go, Chutes Plus and Command Code GOAT list a flat $10 tier. Every vendor price and cap here was read on its own page on 2026-10-04.

Oct 3 · 11 min

Attack chain for the Vibe-Trading advisories: open API on port 8899, file upload and agent shell tools, code running as root in the container, and stolen .env, broker, LLM and cloud keys.

highAI Security Vibe Trading

Vibe-Trading AI Agent Flaws Chain an Open API to Root Code Execution

Three GitHub advisories trace how Vibe-Trading's open API led to a root shell and stolen broker, LLM and cloud keys. Fixed in 0.1.7; upgrade to 0.1.16.

Oct 3 · 8 min

Charcoal poster headed "LightLLM visual nodes: pickle RCE, no fix" with a large 9.8 CVSS v3.1 critical score from VulnCheck, a red "Unpatched, no fix as of 3 Oct 2026" stamp, affected versions up to 1.2.0, CVSS v4.0 9.3, and no reported exploitation not on CISA KEV .

highAI Security Cve 2026 103395

LightLLM Visual Nodes Expose Unauthenticated Pickle RCE (CVE-2026-103395), With No Fix Yet

LightLLM through 1.2.0 runs an unauthenticated, pickle-enabled RPyC service on visual_only nodes. Anyone who can reach the port can run code, and no fix exists yet.

Oct 3 · 6 min

Data graphic: GitLab AI Gateway sandbox escape, CVE-2026-90970, rated CVSS 9.9 Critical. Fixed in 19.2.4, 19.3.2 and 19.4.1 for self-hosted gateways; the attacker must be a logged-in Duo Agent Platform user; no exploitation reported and not in KEV.

highAI Security Cve 2026 90970

GitLab AI Gateway Flaw CVE-2026-90970 Lets Duo Agent Users Run Commands on Self-Hosted Gateways

GitLab fixed a CVSS 9.9 prompt-template sandbox escape in its AI Gateway. Self-hosted gateways must move to 19.2.4, 19.3.2 or 19.4.1.

Oct 2 · 7 min