CVE-2026-79717: Server-Side Request Forgery (SSRF) in Red Hat Red Hat Ansible Automation Platform 2

The vulnerability exists in galaxy_ng, the Ansible Galaxy server plugin for Pulp, where an authenticated user with namespace management permissions can specify a namespace avatar URL pointing to arbitrary addresses, including internal networks, loopback, or cloud instance metadata endpoints. A background worker fetches the avatar URL without validating the destination, allowing the attacker to probe internal services and enumerate reachable IP addresses. The HTTP client used for fetching does not enforce an overall timeout, which can be exploited to pin workers and cause denial of service. The SSRF is blind, as response bodies are discarded unless they resemble images, so secrets from cloud metadata endpoints are not directly exposed. This vulnerability affects Red Hat Ansible Automation Platform versions 2.4 through 2.7 where galaxy_ng is included as Private Automation Hub. Red Hat rates this vulnerability as moderate (medium) severity with a CVSS 3.1 base score of 6.4, reflecting low confidentiality impact, no integrity impact, and low availability impact due to potential denial of service.

An authenticated user with namespace-change permissions can cause Pulp workers to send requests to arbitrary internal or cloud metadata addresses, effectively using the server as a proxy to enumerate internal IP addresses and ports. The SSRF is blind, so attackers do not receive response bodies except for image-like responses, limiting direct data exposure. However, the attacker can cause denial of service by exploiting the lack of HTTP client timeouts, pinning background workers with slow or unresponsive targets. This can disrupt normal operation of the affected service. The vulnerability does not allow direct extraction of cloud metadata secrets but enables internal network reconnaissance and potential denial of service.

There is no complete mitigation or official patch available for this vulnerability at this time. Red Hat recommends the following risk reduction measures: 1) Restrict the galaxy.change_namespace and galaxy.add_namespace permissions to trusted administrators only to limit who can trigger avatar URL downloads. 2) In cloud deployments, protect instance metadata endpoints using hop-limited controls such as AWS IMDSv2 with HttpPutResponseHopLimit=1 to reduce exposure. 3) Apply egress network filtering on Pulp worker nodes to block access to RFC1918 private IP ranges, loopback, link-local, and cloud metadata addresses that are not required for normal operation. Users should monitor Red Hat advisories for any future updates or fixes.