The proxy middleware in mcp-use's inspector forwards requests to a destination the caller names. mountMcpProxy in libraries/typescript/packages/inspector/src/server/proxy/mcp-proxy.ts read the target from the X-Target-URL header or the __mcp_target parameter and proxied to it without inspecting the host, so loopback, link-local and private addresses were all accepted, as were names that resolve to them, and the validation was not reapplied to a redirect the destination returned. A caller could therefore make the server issue requests to addresses reachable only from the host it runs on and read the responses. The current code calls isSafeProxyTarget, which checks the resolved address against private, loopback and link-local ranges before proxying and bounds the number of redirects followed.
CVE-2026-81091
Score 8.6 from GitHub Security Advisory (severity: HIGH) published 2026-08-27. the CNA's CVSS baseline 8.6; sources differ by 0.0.
- High severity, but no confirmed exploitation yet
No vendor fix yet — apply a workaround or compensating control (WAF / firewall / segmentation) and watch for a patch.
- CVSS v3
- 8.6
- EG Score
- 8.6(high)
- EG Risk
- 54(Track)EG Risk 54/100SSVC: Track
EG Risk is EchelonGraph's 0–100 priority score: it fuses intrinsic severity with real-world exploitation and automatability so you can rank equal-severity CVEs and fix the most dangerous first. Higher = act sooner. Distinct from the 0–10 EG Score (severity).
How it’s computedSeverity86% × 45%Exploitation0% × 40%Automatability100% × 15%Action: Routine — remediate on your standard cadence. - EPSS PROB
- 0%
- EPSS %ILE
- 27%
- KEV
- Not listed
Published
August 27, 2026
Last Modified
August 29, 2026
Advisory Details (2)
Auto-updated Aug 27, 2026mcp-use Inspector Proxy Server-Side Request Forgery via Caller-Supplied Target URL | Advisories | VulnCheck
https://www.vulncheck.com/advisories/mcp-use-inspector-proxy-server-side-request-forgery-via-caller-supplied-target-urlGitHub - mcp-use/mcp-use: The fullstack MCP framework to develop MCP Apps for ChatGPT / Claude & MCP Servers for AI Agents. · GitHub
https://github.com/mcp-use/mcp-useVendor Advisories for CVE-2026-81091(1)
These vendors published their own advisory mentioning this CVE — often with vendor-specific remediation steps + affected product lists not in NVD.
Weakness Classification(1)
MITRE Common Weakness Enumeration — the root-cause categories this CVE belongs to.
Data Freshness Timeline
(refreshed 16× in last 7d / 16× in last 30d)
Each row is a source pipeline that fetched or updated this CVE on that date, with what changed. For example, "NVD update" means NVD published or revised its analysis for this CVE; "MITRE cvelistV5" means we ingested or refreshed it from the CNA feed. Most recent first.
- 2026-08-30 01:22 UTCEPSS rescore
- 2026-08-29 23:05 UTCGHSA enrichment
- 2026-08-29 11:59 UTCEG score recompute
- 2026-08-29 11:59 UTCGHSA enrichment
- 2026-08-29 05:51 UTCEG score recompute
- 2026-08-29 05:51 UTCGHSA enrichment
- 2026-08-28 21:42 UTCEPSS rescore
- 2026-08-28 18:47 UTCGHSA enrichment
- 2026-08-28 07:40 UTCGHSA enrichment
- 2026-08-27 20:27 UTCEG score recompute
- 2026-08-27 20:27 UTCGHSA enrichment
- 2026-08-27 20:18 UTCEG score recompute
- 2026-08-27 20:18 UTCGHSA enrichment
- 2026-08-27 17:27 UTCEG score recompute
- 2026-08-27 16:47 UTCEG score recompute
- 2026-08-27 16:43 UTCMITRE cvelistV5first tracked
Frequently asked(5)
What is CVE-2026-81091?
When was CVE-2026-81091 disclosed?
Is CVE-2026-81091 actively exploited?
What is the CVSS score of CVE-2026-81091?
How do I remediate CVE-2026-81091?
Dependency Blast Radius
Explore the affected products and dependency analysis for CVE-2026-81091
Is Your Infrastructure Affected by CVE-2026-81091?
EchelonGraph automatically scans your cloud infrastructure and maps CVE exposure using blast radius analysis.