RabbitMQ implements AMQP 1.0 support via an optimized server‑mode parser that matches symbolic data, sequence, and value body descriptors by textual prefixes. Due to a logic flaw, the parser does not verify that the encoded symbol lengths match the textual prefix. An authenticated AMQP 1.0 publisher with write permission can therefore send a well‑formed but unknown descriptor that begins with a registered body descriptor. The broker misclassifies the incoming data, stores the bytes as a standard body section, and, when a strict AMQP 1.0 consumer later attempts to decode the entry, the consumer process terminates. If the data is converted to AMQP 0‑9‑1 or another cross‑protocol format, full body decoding can also terminate the consumer, allowing the attacker to repeatedly cause failures through acknowledgement‑required delivery. The broker itself remains alive and routing and authorization are not affected.

This flaw exists in RabbitMQ server versions 4.0.0 through 4.0.24, 4.1.15, 4.2.10, and 4.3.4. All earlier releases not mentioned are not affected, and the issue is fixed in the listed releases.

With a CVSS score of 5.3 the vulnerability is of moderate severity. EPSS indicates a very low exploitation probability of 0.54%, and the flaw is not listed in the CISA KEV catalog, indicating no widespread exploitation at the time of analysis. Exploit requires authentication and write permission to publish AMQP 1.0 messages, so an attacker who has compromised credentials or misconfigured permissions can trigger consumer crashes. Because the broker remains operational and routing is unaffected, the impact is limited to service availability for consumers rather than data loss or integrity compromise. The risk therefore depends on the presence of write‑enabled AMQP 1.0 publishers and the criticality of consumers in the environment.