Symfony Mailomat Mailer CVE-2026-48747

Exploitation requires that the attacker can send HTTP POST requests to the Symfony application's Mailomat webhook route - a network-accessible endpoint configured to receive Mailomat callbacks, typically with no authentication gatekeeping (webhook receivers are public by design). The Mailomat Mailer bridge (symfony/mailomat-mailer) must be installed and a webhook listener route must be registered in the application. The attacker must additionally be able to construct a valid HMAC signature for the chosen weak algorithm without knowing the secret; for HMAC-MD4, published existential forgery techniques (Contini & Yin, ASIACRYPT 2006) satisfy this requirement. For algorithms where no such forgery is currently known (e.g., HMAC-SHA512), exploitation would require knowing the secret and the vulnerability would be moot. Exploitation is NOT possible against deployments that do not use the Mailomat Mailer bridge or have not registered a webhook endpoint. Exploitation is NOT possible if the endpoint is network-isolated or the X-MOM-Webhook-Signature header is validated externally (e.g., by a WAF) before reaching Symfony.

No CVSS vector or EPSS score is present in the input data; all metric assessments below are independently derived. The webhook endpoint is network-accessible and unauthenticated (webhook receivers are public by design), establishing a network attack vector with no privilege requirement. However, practical exploitation currently requires the attacker to exploit a cryptographic weakness in a specific HMAC primitive - for example, HMAC-MD4 existential forgery as published by Contini and Yin (ASIACRYPT 2006) - without knowing the shared secret, placing this in the high-complexity tier. The vulnerability is not yet in CISA KEV, and no public proof-of-concept exploit code has been identified, both confirmed by input data. The real-world impact is limited to integrity of webhook events (fraudulent delivery, bounce, or complaint signals injected into the application) with no direct confidentiality or availability impact. The forward-looking risk is notable: the attack surface expands with every new HMAC primitive PHP exposes, and any future cryptanalytic advance in any accepted primitive is immediately exploitable without a code change on the Symfony side. The tags Information Disclosure and JWT Attack confirm the algorithm-confusion classification. Overall this is a genuine architectural flaw with moderate current exploitability and a growing risk profile over time, warranting prompt patching rather than emergency response.

An attacker with network access to a Symfony application's Mailomat webhook endpoint crafts a malicious POST payload simulating a fraudulent email delivery event, then computes its HMAC using the HMAC-MD4 forgery technique (Contini & Yin, ASIACRYPT 2006) without knowledge of the webhook secret. The attacker sets the X-MOM-Webhook-Signature header to md4=<forged_value>; the unpatched validateSignature() method accepts the attacker-supplied algorithm, calls hash_hmac('md4', ..., $secret), and the constant-time comparison passes against the forged value. The application processes the fabricated event as an authentic Mailomat callback, potentially triggering downstream logic such as marking emails delivered, suppressing retry queues, or updating user-facing delivery status.