/

CVE-2026-74905

Severity:

HIGH RISK

(7.1)

Server-Side Request Forgery (CWE-918) in SiYuan Proxy

CVE-2026-74905

Deferred

Vulnerability type

Category:

Server-Side Request Forgery

CWE:

CWE-918

Impact:

Internal Service Access and Cloud Credential Exposure

Severity:

HIGH RISK

(7.1)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

SiYuan contains an SSRF protection bypass in SafeMode because its private-address check does not recognize IPv6 transition addresses that can represent private IPv4 destinations. An authenticated attacker may use proxy endpoints to access internal services, cloud metadata endpoints, and returnable response content when the application can resolve and connect to attacker-supplied destinations.

Affected context

The vulnerability occurs in SiYuan deployments that enable SafeMode and expose the network forward proxy, WebSocket proxy, or SSE proxy functionality to authenticated users. The risk is present when hostname resolution produces IPv6 transition addresses, such as NAT64, 6to4, or Teredo addresses, that embed or route to otherwise restricted private IPv4 destinations.

Exploit preconditions

  • Attack surface: SiYuan network proxy endpoints
  • Required access: Valid authenticated access to a feature that accepts a destination URL
  • Trigger point: URL resolution and outbound connection through the proxy
  • Why exploitation is possible: The SSRF guard checks common private, loopback, link-local, and unspecified ranges but does not adequately classify IPv6 transition addresses that can reach private IPv4 targets

Why this happens

The underlying bug is incomplete destination-address validation in the isPrivateIP function used by the SSRF-safe dialer. The check recognizes several directly represented restricted address classes but omits IPv6 transition mechanisms that encode or route to private IPv4 destinations. This maps to CWE-918: Server-Side Request Forgery, where attacker-controlled network destinations are accepted without sufficiently comprehensive validation.

Potential impact

  • Internal services could be accessed from the SiYuan server
  • Cloud instance metadata and temporary credentials could be exposed
  • Internal infrastructure could be enumerated through response behavior
  • Sensitive response bodies could be returned to the attacker
  • Network segmentation and egress restrictions could be bypassed
  • Compromised credentials could enable further access or lateral movement

General mitigation guidance

Mitigation generally focuses on making outbound destination validation comprehensive and reducing the trust placed in user-controlled URLs.

Common defensive patterns include:

  • Resolve hostnames before connection and validate every resulting address
  • Treat IPv4-mapped, NAT64, 6to4, Teredo, and other transition addresses according to their effective destination
  • Revalidate the final socket destination to reduce DNS rebinding and resolution inconsistencies
  • Prefer explicit outbound allowlists over broad URL acceptance
  • Restrict proxy functionality to trusted administrative workflows where possible
  • Block access to cloud metadata services and sensitive management ranges at the network layer
  • Apply least-privilege network policies and limit the process's access to internal services

How teams detect issues like this

Organizations can identify similar SSRF risks through:

  • Static analysis for user-controlled URLs passed to HTTP clients, dialers, or proxy handlers
  • Code review checks for IP filtering that validates only textual or incomplete address categories
  • Tests covering IPv4, IPv6, mapped addresses, DNS rebinding, NAT64, 6to4, and Teredo representations
  • Egress monitoring for application connections to private ranges, metadata endpoints, and unexpected internal ports
  • Policy enforcement requiring outbound allowlists for server-side fetch functionality
  • Neutral automated review tools that flag network requests using attacker-influenced destinations

Sources

SourceLink
github.comgithub.com/siyuan-note/siyuan/security/advisories/GHSA-qq8m-8p8v-x4xg
vulncheck.comvulncheck.com/advisories/siyuan-before-ssrf-via-ipv6-transition-address-bypass

Frequently asked

questions

It may be exploitable over the network when an attacker has authenticated access to an exposed SiYuan proxy endpoint and the deployment can make outbound connections.

The described attack requires authentication and access to a proxy feature that accepts a user-controlled destination URL. The exact scope depends on the deployment's endpoint exposure and authorization settings.

Use comprehensive destination validation after resolution, account for IPv6 transition mechanisms, and combine application-layer checks with network-level egress restrictions.

Minimize server-side fetching, use explicit destination allowlists, test all address representations, protect metadata endpoints, and apply least-privilege network policies.

Find out what's already
exploitable in your codebase.