CVE-2026-57585: MessagePack Out-of-Bounds Read / DoS Vulnerability

Technical summary

CVE-2026-57585 is an out-of-bounds read vulnerability in the MessagePack Python library stemming from improper state management in the Unpacker class. When deserialization fails and raises an exception, the Unpacker's internal buffer pointers and state flags are not correctly reset. Subsequent calls to unpack data using the same Unpacker instance cause reads beyond allocated memory boundaries, triggering a segmentation fault. This maps to CWE-416 (Use After Free), as the underlying mechanism involves accessing freed or deallocated memory regions within the deserialization context. The issue manifests only when the Unpacker is reused after catching an exception, making certain application patterns more vulnerable than others.

Business impact

Applications relying on MessagePack for data deserialization—including those handling API responses, distributed system communication, or streaming protocols—face availability risks. A remote attacker can craft malformed MessagePack payloads that, when processed by a vulnerable Unpacker instance in a loop or event handler, cause application crashes. This enables denial-of-service attacks without requiring authentication or special privileges. For services that automatically retry failed deserializations or batch-process untrusted data, exploitation requires minimal effort and can disrupt availability with moderate operational impact.

Affected systems

The vulnerability affects msgpack-python (the official MessagePack library for Python) in all versions prior to 1.2.1. Any application using this library to deserialize untrusted data is potentially affected. Common use cases include web services consuming JSON-serialized alternatives or protocol handlers in microservices architectures, message brokers, and data processing pipelines. The impact is amplified in long-running processes (daemons, services) where the same Unpacker object is reused across multiple requests or messages.

Exploitability

Exploitation requires only network access to transmit a malformed MessagePack payload to a vulnerable application. No authentication, user interaction, or special system privileges are required. The CVSS score of 7.5 (HIGH) reflects the ease of triggering a denial-of-service condition via unauthenticated network input. However, successful exploitation depends on the target application's deserialization pattern: services that create a fresh Unpacker per message are not vulnerable, while those that reuse a single instance across multiple messages are at risk. Proof-of-concept development is straightforward for an attacker with basic MessagePack format knowledge.

Remediation

Upgrade msgpack-python to version 1.2.1 or later. This release includes fixes for the Unpacker state initialization issue, ensuring that error conditions properly reset internal buffers and flags. Additionally, applications may reduce exposure by creating a new Unpacker instance for each message or by implementing per-message deserialization scope rather than maintaining a long-lived Unpacker object. Organizations should prioritize patching production systems, particularly those handling untrusted or adversary-controlled MessagePack data.

Patch guidance

Consult the official msgpack-python release notes to confirm the patched version for your environment. Version 1.2.1 and later are safe. Update via your standard package management tool (pip, conda, or distribution repositories). For organizations using vendored or bundled versions of MessagePack, verify that the package version has been updated to 1.2.1 or later. Test the patched version in a staging environment to confirm compatibility with your application's deserialization workflows before deploying to production.

Detection guidance

Monitor application logs and system error logs for segmentation faults (SIGSEGV) or process crashes occurring immediately after MessagePack deserialization errors. Use intrusion detection or network monitoring to identify suspicious MessagePack payloads that deviate from normal schema patterns. Implement application-level logging to track Unpacker reuse patterns: log when an Unpacker instance is created, when errors occur during unpacking, and when the same instance is reused afterward. In development and testing, use sanitizers (AddressSanitizer, MemorySanitizer) to detect out-of-bounds memory access before deployment.

Why prioritize this

This vulnerability merits prompt patching due to its combination of high accessibility (unauthenticated network attack vector), ease of exploitation, and denial-of-service impact. Services that are performance-critical or mission-essential and rely on MessagePack for data handling should be prioritized. Additionally, any application exposing a MessagePack endpoint to untrusted or public networks is at elevated risk. While not as critical as memory corruption vulnerabilities that enable arbitrary code execution, the ability to trigger reliable crashes with minimal effort and no prerequisites makes this a significant availability threat.

Risk score, explained

The CVSS 3.1 score of 7.5 reflects a HIGH severity vulnerability. The vector CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H breaks down as: network-accessible (AV:N), low attack complexity (AC:L), no privileges required (PR:N), no user interaction (UI:N), impact confined to the affected system (S:U), and high availability impact (A:H). The absence of confidentiality or integrity impact (C:N, I:N) reflects that crashes do not leak data or corrupt application state; however, the trivial exploitability and guaranteed availability disruption justify the HIGH classification.

Frequently asked questions