Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.
Sources: National Vulnerability Database, CISA KEV, Exploit-DB
CVE-2026-85738 Overview
CVE-2026-85738 is a Server-Side Request Forgery (SSRF) vulnerability in TREK, a collaborative travel planner application. The flaw resides in the checkSsrf logic within server/src/utils/ssrfGuard.ts. The guard fails to recognize IPv6 transition address formats (NAT64, 6to4, and Teredo) that embed IPv4 destinations. An authenticated attacker who controls a DNS record can return an AAAA response pointing to a transition address that encodes a private, loopback, or link-local IPv4 target. The isAlwaysBlocked and isPrivateNetwork checks classify these addresses as allowed, enabling egress to internal services or cloud metadata endpoints. The issue is fixed in version 3.4.0.
Critical Impact
Authenticated attackers can bypass SSRF protections to reach internal services or cloud metadata endpoints via user-facing imports, webhooks, integrations, or plugin egress, exposing returned data.
Affected Products
- TREK collaborative travel planner, all versions prior to 3.4.0
- Deployments that route NAT64, 6to4, or Teredo IPv6 transition traffic
- Instances exposing imports, webhooks, integrations, or plugin egress to authenticated users
Discovery Timeline
- 2026-09-24 - CVE-2026-85738 published to NVD
- 2026-09-28 - Last updated in NVD database
Technical Details for CVE-2026-85738
Vulnerability Analysis
The vulnerability is classified under [CWE-918] Server-Side Request Forgery. TREK's SSRF guard inspects destination addresses to block requests to internal network ranges. However, the classification logic in isAlwaysBlocked and isPrivateNetwork only accounts for native IPv4 and native IPv6 ranges. IPv6 transition mechanisms encode IPv4 addresses inside IPv6 containers, and these embedded destinations are not decoded before the allow or block decision runs. As a result, an attacker-controlled DNS AAAA record can direct a server-side fetch to a transition address that resolves on the wire to a private IPv4 target such as 127.0.0.1, 169.254.169.254, or an RFC 1918 host.
Root Cause
The root cause is incomplete address classification in server/src/utils/ssrfGuard.ts. The guard does not parse the embedded IPv4 payload of NAT64 (64:ff9b::/96), 6to4 (2002::/16), or Teredo (2001::/32) addresses before evaluating whether the target is private, loopback, or link-local. The checks treat the full IPv6 string as a public address.
Attack Vector
Exploitation requires an authenticated user and a deployment that routes the relevant transition format. The attacker registers a hostname whose AAAA record returns a transition address, for example 2002:7f00:0001:: (6to4 encoding of 127.0.0.1) or 64:ff9b::a9fe:a9fe (NAT64 encoding of 169.254.169.254). The attacker then supplies that hostname to a TREK feature that performs server-side fetches, such as imports, webhooks, integrations, or plugin egress. The SSRF guard allows the request, and the server contacts the embedded internal IPv4 destination. Response data can expose cloud metadata credentials or internal service contents.
Technical references for the fix are available in the GitHub Security Advisory GHSA-9hpq-mrpx-mcxg and the patch commit.
Detection Methods for CVE-2026-85738
Indicators of Compromise
- Outbound DNS AAAA lookups from the TREK server resolving to prefixes 64:ff9b::/96, 2002::/16, or 2001::/32
- Server-side HTTP requests to IPv6 transition addresses in application or proxy logs
- Access attempts to cloud metadata endpoints (169.254.169.254, fd00:ec2::254) originating from the TREK server process
- Unusual webhook, import, or plugin requests submitted by authenticated users containing non-standard IPv6 hostnames
Detection Strategies
- Inspect egress proxy and firewall logs for connections from the TREK server to addresses matching NAT64, 6to4, or Teredo prefixes
- Correlate authenticated user actions (imports, webhook creation, integration setup) with outbound requests to IPv6 transition ranges
- Alert on any server-side fetch whose resolved destination maps to a private IPv4 range after transition decoding
Monitoring Recommendations
- Enable DNS query logging on resolvers used by the TREK server and flag AAAA responses in transition prefixes
- Monitor cloud metadata service (IMDS) access patterns from application workloads and alert on unexpected callers
- Review authentication logs for newly registered or low-reputation accounts creating webhooks or integrations immediately before anomalous egress
How to Mitigate CVE-2026-85738
Immediate Actions Required
- Upgrade TREK to version 3.4.0 or later, which contains the fix. See the v3.4.0 release notes.
- Restrict egress from the TREK server to only required destinations using network policy or an egress proxy allowlist
- Enforce IMDSv2 or equivalent session-token metadata protection on cloud instances hosting TREK
- Audit webhook, integration, and import configurations for suspicious hostnames created before the upgrade
Patch Information
The fix is delivered in TREK 3.4.0 via pull request #1540 and commit a2bd9be. The patched ssrfGuard.ts recognizes NAT64, 6to4, and Teredo transition addresses and decodes the embedded IPv4 destination before applying isAlwaysBlocked and isPrivateNetwork classification.
Workarounds
- Disable IPv6 on the TREK server if transition routing is not required for operation
- Configure the host or container to drop outbound traffic to 64:ff9b::/96, 2002::/16, and 2001::/32 at the network layer
- Place server-side fetch functionality behind an egress proxy that performs its own transition-aware SSRF validation
- Limit webhook, import, integration, and plugin features to trusted user roles until the upgrade is applied
# Example nftables rules to block IPv6 transition prefixes from the TREK host
nft add rule inet filter output ip6 daddr 64:ff9b::/96 drop
nft add rule inet filter output ip6 daddr 2002::/16 drop
nft add rule inet filter output ip6 daddr 2001::/32 drop
Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.
Sources: National Vulnerability Database, CISA KEV, Exploit-DB

