Finding and attribution
An undersized AMQP frame can underflow the computed length and cause an out-of-bounds read during frame parsing. Trail of Bits independently confirmed the finding; Anthropic CVD notes static analysis without a trigger input.
Product: rabbitmq-c. AI system: Claude Mythos Preview. Team / organization: Anthropic / Trail of Bits.
CVE identifiers: CVE-2026-44235.
AI and human contribution (source account): Anthropic CVD explicitly attributes discovery to Claude; Trail of Bits true-positive triage.
Human involvement is only described where the source account establishes it; unknown involvement does not mean none. Vendor confirmation of a flaw does not confirm every claim about AI contribution.
Discovery context: Public vulnerability and AI-discovery attribution are documented. Sources do not establish exploitation in the wild or zero-day status.
Remediation status: fix documented. The vendor advisory documents the fixed rabbitmq-c release; use the latest project release containing this fix.
A source-documented fix does not establish installed patches or current safety. Source checking does not independently reproduce the finding or establish observed exploitation.
Disclosed 17 Sep 2026. AI attribution dated 17 Aug 2026. Sources checked 14 Nov 2026; these are separate dates, not a live installation check.
Recorded CVSS: Medium · highest recorded score 6.5. Documented technical effects: denial of service.
Severity, consequences and exploitation evidence
Case grouping uses the highest recorded CVSS score, with its version-specific severity band. It is descriptive, not a combined or editorial risk index. All recorded assessments remain below; scores from different versions and sources may disagree. CVSS specifications define version-specific scoring and severity bands.
- CVE-2026-44235: CVSS 3.1 · 6.5 (Medium) ·
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:N/I:N/A:H. GitHub_M (CNA; CVE Program record) · checked 16 Nov 2026.
CVE-2026-44235: denial of service.
Prerequisites / affected scope: The CNA description states: rabbitmq-c is a C-language AMQP client library for RabbitMQ. Prior to 0.16.0, a malicious AMQP server can send an undersized HEADER or METHOD frame during client login and cause unsigned size_t underflow in amqp_handle_input() in librabbitmq/amqp_connection.c. The parser subtracts HEADER_SIZE, fixed per-frame fields, and FOOTER_SIZE from state->target_size without first checking the minimum frame length. The wrapped encoded.len value is passed through amqp_decode_properties() to amqp_decode_table_internal(), where it defeats bounds checks and causes an out-of-bounds read and process crash. An on-path attacker can also trigger the issue when AMQP traffic is not protected by TLS with certificate validation. The demonstrated impact is denial of service, with no reliable memory disclosure or code execution shown. This issue is fixed in version 0.16.0..
Exploitation evidence: Exploitation in the wild is not established by the cited description.. Impact source · checked 16 Nov 2026.
Documented fix: The vendor advisory documents the fixed rabbitmq-c release; use the latest project release containing this fix.