/

CVE-2026-73629

Severity:

HIGH RISK

(8.5)

Server-Side Request Forgery (CWE-918) in Serendipity

CVE-2026-73629

Deferred

Vulnerability type

Category:

Server-Side Request Forgery

CWE:

CWE-918

Impact:

Internal service access and sensitive response disclosure

Severity:

HIGH RISK

(8.5)

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

Serendipity contains a server-side request forgery vulnerability in the

serendipity_url_allowed() filter. An authenticated user with the

adminImagesAdd permission may use alternate IP address representations to

bypass filtering, request internal services, and retrieve response bodies

through the public uploads directory.

Affected context

The issue occurs in Serendipity deployments where the URL allowlisting filter

processes user-controlled image or upload-related URLs. Risk is increased when

authenticated users have the adminImagesAdd permission and the server can

reach internal networks, IPv6 endpoints, or link-local services. The affected

validation logic may not recognize hexadecimal IPv4 representations, IPv6

literals, or link-local address ranges as internal destinations.

Exploit preconditions

  • Attack surface: The image or upload URL handling functionality protected

by the adminImagesAdd permission

  • Required access: An authenticated account with adminImagesAdd
  • Trigger point: Processing a user-supplied URL through the

serendipity_url_allowed() filter

  • Why exploitation is possible: Address validation does not consistently

canonicalize or reject alternate representations of internal IPv4, IPv6,

and link-local destinations

  • Disclosure path: Retrieved response content may become accessible through

the public uploads directory

Why this happens

The vulnerability is caused by incomplete server-side URL validation and

address canonicalization. The filtering logic checks destination formats

without reliably normalizing hexadecimal IPv4 addresses, IPv6 literals, and

link-local ranges before applying network-boundary rules. This allows a

user-controlled URL to resolve to an internal destination that the filter

intended to block, mapping to CWE-918: Server-Side Request Forgery.

Potential impact

  • An attacker may access services on internal network interfaces.
  • Sensitive response bodies could be retrieved through a publicly reachable

uploads path.

  • Cloud or infrastructure metadata may be exposed where reachable.
  • Internal administrative APIs may be probed or accessed.
  • The server could be used as a pivot toward other systems.

General mitigation guidance

Mitigation generally focuses on enforcing destination validation after full

URL and address normalization.

Common defensive patterns include:

  • Canonicalizing hostnames and IP addresses before applying allowlists or

denylists.

  • Rejecting private, loopback, link-local, multicast, unspecified, and

other non-public address ranges across IPv4 and IPv6.

  • Resolving hostnames and validating every resulting address, including

protections against DNS rebinding.

  • Using an explicit outbound destination allowlist rather than relying only

on blocklists.

  • Preventing fetched response bodies from being written to publicly

accessible locations unless strictly required.

  • Restricting outbound network access and running the web process with

minimal privileges.

How teams detect issues like this

Detection approaches can include:

  • Static analysis and code review rules for URL fetches that lack

canonicalization and post-resolution network checks.

  • Review signals such as direct string comparisons against IP addresses,

incomplete IPv4-only validation, and unrestricted redirects.

  • Dynamic tests covering hexadecimal IPv4 notation, IPv6 literals,

link-local targets, redirects, DNS rebinding, and encoded host forms.

  • Network monitoring for unexpected web-server requests to private,

loopback, metadata, or link-local address ranges.

  • Policy enforcement that restricts outbound connections from application

workloads and prevents sensitive fetched content from entering public

asset directories.

Sources

SourceLink
github.comgithub.com/s9y/Serendipity/security/advisories/GHSA-2m48-gjj5-5x86
vulncheck.comvulncheck.com/advisories/serendipity-before-ssrf-via-hex-ipv4-and-ipv6-addresses

Frequently asked

questions

It can be exploited through the Serendipity web application by an

authenticated user who has the required image-management permission. The

attacker does not necessarily need direct network access to the internal

destination.

The relevant account must have the adminImagesAdd permission. The server

must also be able to connect to the targeted internal or link-local service.

Normalize and resolve destinations before validation, reject all non-public

address ranges across IPv4 and IPv6, and use explicit outbound allowlists.

Public exposure of fetched response bodies should also be avoided.

Combine centralized SSRF-safe URL handling with outbound network controls,

redirect and DNS-rebinding protections, least-privilege service accounts,

and automated tests for alternate address representations.

Find out what's already
exploitable in your codebase.