Agent Security & SandboxingOpenClaw / depthfirst (Mav Levin)OpenClaw

OpenClaw ships a fix for CVE-2026-25253

Made by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made

Shipped a fix for CVE-2026-25253 — the Control UI accepted unauthenticated WebSocket connections and trusted gatewayUrl without validation, enabling token theft followed by one-click RCE (CVSS 8.8)

Why it matters

The first major incident to shatter the assumption that local agents are safe. Origin validation has since become a baseline requirement

Source — Primary sources first; secondary reports are labelled “Press”

Pressthehackernews.comhttps://thehackernews.com/2026/02/openclaw-bug-enables-one-click-remote.html

Added in Week 35, 2026 · See that week

On the map — ㉒ Agent Security line

  1. Previous stopCisco finds 26% of 31,000 agent skills have at least one vulnerability
  2. This recordOpenClaw CVE
  3. Next stopDefenseClaw
See it on the map

Same week

OpenClaw name settled (Warelay→Clawdbot→Moltbot→OpenClaw)PressMoltbot is renamed OpenClaw and launches an official blogOfficialMoonshot AI releases Kimi K2.5OfficialMETR releases Time Horizon 1.1Official

Nearby on the same axis

Cisco finds 26% of 31,000 agent skills have at least one vulnerabilityOfficialCisco releases DefenseClawOfficialAnthropic reports prompt injection defense results (Claude for Chrome)OfficialOpenClaw publishes its security roadmapOfficial