LodeHQ
AI news for people who shipFree, by email

Infosec · Wednesday, 29 July 2026 · Issue 58 · 7 stories

Ruflo flaw hijacks AI agents, TeamCity RCE, 24K BMCs

Vulnerabilities & Exploits

CVE-2013-4786 Exposes 24K+ Internet-Facing BMCs, Enabling Mass Offline Password Cracking

LavaHQ scanned UDP port 623 and identified 36,872 internet‑exposed IPMI interfaces. A flaw (CVE‑2013‑4786) caused 24,650 of them to reveal password‑derived HMAC‑SHA1 hashes, making bulk offline cracking trivial; dozens of weak, factory defaults were recovered, confirming active exploitation. A compromised BMC gives attackers full control over server infrastructure.

lavahq.io ↗13 min read at source

Critical Ruflo Flaw Lets Attackers Hijack AI Agents and Steal API Keys

A CVSS 10.0 vulnerability (CVE‑2026‑59726) in Ruflo’s MCP bridge exposed unauthenticated POST /mcp endpoints, letting attackers execute commands, obtain a shell in the bridge container, read provider API keys, and poison the AgentDB learning store. All Ruflo releases before 3.16.3 are vulnerable; the issue is patched in version 3.16.3.

nvd.nist.gov ↗1 min read at source

Threats & Malware

Tengu botnet hijacks Linux watchdog to reboot after process kill

Nozomi Networks discovered Tengu, a Mirai‑derived IoT botnet that triggers a hardware watchdog reboot when defenders terminate its main process. The forced reboot reactivates its hidden guardian and other persistence hooks, letting the malware regain control on compromised Linux devices. The threat spreads via Telnet credential brute‑force and supports dozens of DDoS attacks.

nozominetworks.com ↗11 min read at source

Privacy, Policy & Governance

You’re caught up.

That was all 7 stories in the 29 July issue.