Red Hat AWX CVE-2026-71365

Exploitation requires three concurrent conditions: (1) the attacker must hold the admin role on a specific AWX job template - this is an authenticated, low-privilege requirement (CVSS PR:L) but limited to template-level administration, not system-wide AWX admin; (2) that job template must have GitHub webhook integration enabled with a signing key configured, since the signing key is the prerequisite for forging a valid HMAC-signed payload; and (3) a Git Personal Access Token must be configured on the job template, as without a PAT there are no credentials to exfiltrate. Limiting factors: anonymous or unauthenticated external attackers cannot exploit this; exploitation is gated by obtaining template admin access first. The attack is not automatable at scale across random targets because it requires template-specific secrets. SSRF impact against internal services is further constrained by network topology - only hosts reachable from the AWX server are in scope.

The CVSS 7.7 vector (AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:N/A:N) accurately models the core threat: network-reachable, low complexity, requiring only job-template admin privileges, with a Changed scope that elevates impact beyond the AWX instance itself. The C:H rating is grounded in Git PAT exfiltration, which in practice grants the attacker read or write access to source repositories - a significant blast radius in CI/CD environments where AWX orchestrates deployments. EPSS data is not provided; given no KEV listing and no public POC, opportunistic mass exploitation is unlikely. Real-world risk is highest in organizations where AWX job-template admin is granted broadly or where a single PAT is reused across many repositories. The PR:L requirement meaningfully limits exposure: an anonymous external attacker cannot exploit this without first obtaining template admin access. The most credible threat model is an insider or a compromised low-privileged account escalating impact by stealing Git credentials. SSRF to internal network endpoints is a secondary risk - severity depends on what services are reachable from the AWX host.

An attacker who has obtained job-template admin privileges on an AWX instance - whether through legitimate assignment, credential theft, or account compromise - navigates to a webhook-enabled job template, reads the GitHub webhook signing key, and crafts a forged GitHub pull request webhook JSON payload in which pull_request.statuses_url points to an attacker-controlled HTTPS endpoint. The attacker computes a valid HMAC-SHA256 signature over the payload using the stolen key and submits the webhook to AWX's public webhook receiver. AWX validates the signature, persists the malicious statuses_url in job extra variables, and when the job completes, POSTs a status update to the attacker's server with the Git PAT in the Authorization header, fully disclosing the credential. No public POC exists at time of analysis, but the attack mechanics are straightforward given the description.