/

CVE-2026-54491

Severity:

HIGH RISK

(7.1)

SSRF (CWE-918) in Koel Podcast and Radio Fetch Paths

CVE-2026-54491

Deferred

Vulnerability type

Category:

Server-Side Request Forgery

CWE:

CWE-918

Impact:

Internal service access and unintended response disclosure

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

Koel contains server-side request forgery weaknesses in podcast and internet radio fetching paths because validated destinations are not consistently pinned or rechecked across DNS resolution and redirects. An authenticated user who can submit or refresh a remote media URL may cause the server to contact internal services, cloud metadata endpoints, or other unintended destinations and may receive parsed or streamed response content.

Affected context

The issue may occur in Koel deployments that expose podcast or internet radio APIs and use outbound URL-fetching functionality. Risk is higher when an authenticated user can control, submit, or refresh remote URLs and when the server can reach private network services, cloud metadata endpoints, or other restricted destinations. Relevant processing includes podcast creation and refresh, radio station creation, stream URL handling, content-type validation, and obsolete-podcast checks.

Exploit preconditions

  • Attack surface: Koel podcast and internet radio API flows that fetch or validate user-supplied URLs
  • Required access: An authenticated account with permission to create, update, or refresh podcast or radio resources
  • Trigger point: Server-side URL resolution, redirect following, stream retrieval, or audio content-type validation
  • Why exploitation is possible: A destination may pass an initial public-host check but resolve differently later, or a redirect may lead to an internal address without equivalent validation at every hop

Why this happens

The vulnerability is caused by incomplete server-side request validation. A public-host or safe-URL decision is made at one point in time, but the validated network address is not consistently pinned and redirect targets are not uniformly revalidated.

This creates a trust-boundary failure in which attacker-influenced URL behavior, DNS changes, or redirects can cause the server to connect to destinations that were not approved by the original check. The behavior maps to CWE-918: Server-Side Request Forgery, where server-side network access can be redirected toward unintended resources.

Potential impact

  • An attacker may access internal HTTP services that are unreachable from the public internet
  • The server could contact cloud instance metadata endpoints
  • Sensitive response content may be disclosed through parsed or streamed results
  • Internal services could be probed or abused through the Koel application
  • Network segmentation and access-control assumptions could be weakened

General mitigation guidance

Mitigation generally focuses on making outbound URL access safe by default and preserving validation decisions throughout the entire request lifecycle.

Common defensive patterns include:

  • Resolving hostnames and validating every resulting address against approved network ranges
  • Pinning the validated destination or otherwise preventing DNS changes between validation and connection
  • Revalidating every redirect target before following it
  • Restricting outbound requests to an explicit protocol, port, hostname, and destination allowlist
  • Disabling redirects when they are not required for the feature
  • Limiting response size, connection duration, and accessible network paths
  • Running fetch workers with minimal network privileges and no unnecessary access to metadata services
  • Separating media retrieval from sensitive application and control-plane networks

How teams detect issues like this

High-level detection can combine code inspection, runtime monitoring, and policy enforcement.

Useful approaches include:

  • Static analysis searches for URL validation that is separated from the eventual socket connection
  • Code review checks for hostname checks without address pinning, redirect handling without per-hop validation, and DNS lookups performed more than once
  • Tests covering redirects to loopback, link-local, private, IPv6-local, and reserved address ranges
  • Outbound network policy alerts for application workers contacting internal ranges or cloud metadata endpoints
  • Logging of requested URLs, resolved addresses, redirect chains, and blocked destinations without recording sensitive response bodies
  • Automated security review tools that identify SSRF-prone HTTP client and URL-fetching patterns

Sources

SourceLink
github.comgithub.com/koel/koel/commit/5f6ce2cefd08f437a269236b677ad971517ccbb6
github.comgithub.com/koel/koel/commit/c264a3d52513a83b21e1cc3a20e895caea97fc4a
github.comgithub.com/koel/koel/pull/2546
github.comgithub.com/koel/koel/pull/2549
github.comgithub.com/koel/koel/releases/tag/v9.7.1
github.comgithub.com/koel/koel/security/advisories/GHSA-6qvr-wjmv-v8mm

Frequently asked

questions

It may be exploitable through Koel's network-accessible podcast and radio APIs, but an attacker generally needs an authenticated account with permission to submit or refresh remote media URLs.

The relevant account must be able to reach a podcast or internet radio operation that causes Koel to validate, follow, or fetch an attacker-influenced URL. The practical impact also depends on the server's network connectivity and permissions.

Apply strict outbound URL allowlisting, validate every redirect and resolved address, and ensure the address approved during validation is the one used for the connection.

Treat all user-influenced URLs as untrusted, restrict protocols and destinations, disable unnecessary redirects, apply egress network controls, and test DNS rebinding and redirect behavior as part of security review.

Find out what's already
exploitable in your codebase.