Scans for LiteLLM/Starlette (CVE-2026-42271, CVE-2026-48710), (Mon, Nov 16th)
Honeypot detects active scans exploiting chained LiteLLM/Starlette CVEs for unauthenticated RCE.
Security researchers observed weekend attacks against honeypots exploiting CVE-2026-42271 and CVE-2026-48710 in Starlette and LiteLLM. The vulnerabilities stem from Starlette versions prior to 1.0.1 failing to validate the Host header, allowing attackers to bypass access controls on the /mcp-rest/test/tools/list endpoint and achieve remote code execution without authentication. Attackers are modifying attack payloads with unusual characters like brackets to evade existing WAF filters.
AI summary of the linked source. Not reviewed by an editor. How this works
Run Starlette?
Get an email when a new story names Starlette, usually within the hour.
Free. Your list stays private and never appears in a subject line. One click stops it. How Vendor Watch worksPrivacy
The lesson from this incident
Full lessonAttackers actively exploited CVE-2026-42271 and CVE-2026-48710 within days of disclosure, targeting Starlette versions prior to 1.0.1 that fail to validate the Host header. This oversight allowed adversaries to bypass access controls on a sensitive endpoint and achieve remote code execution without any authentication. The speed of exploitation underscores the critical importance of patching internet-facing AI/LLM infrastructure promptly. Notably, attackers modified payloads with unusual characters (e.g., brackets) to evade WAF signatures, demonstrating that WAF rules alone are insufficient without underlying patches. Organizations running LiteLLM or any Starlette-based service must treat this as a high-severity, time-sensitive patching event.
What helps against this
How we list vendors- Vulnerability scanning and prioritisation
You know which systems are vulnerable, and which ones to fix first.
- Automated patch management
Known holes get closed in days, not months.
- Web application firewall and DDoS protection
Websites and APIs stay up, and common web attacks are filtered before they reach them.
Indicators of Compromise
Addresses and file hashes are copied from the publisher's page by pattern matching. Names are picked by AI. On stories from before 17 November 2026, some addresses and hashes were picked by AI too. Not reviewed by an editor. How this works
- cve: CVE-2026-48710
- cve: CVE-2026-42271