cve.bz

CVE-2026-106439 hydra-ecosystem hydra: Reflection Injection

8.5 highEPSS 0.5%Fix available

High (8.5) in hydra-ecosystem hydra. EPSS 0.5% estimated probability of exploitation activity in the next 30 days. Fixed in hydra-core 1.3.7, 1.4.0.dev10.

On this page
Hydra is a framework for elegantly configuring complex applications. From 1.3.4 until 1.3.7 and 1.4.0.dev10, Hydra stores legacy instantiate target blocklists and related execution-policy collections in mutable module-level state. An attacker who controls multiple sibling target entries can resolve hydra._internal.target_policy.UNCONTROLLED_EXECUTION_TARGETS.discard through instantiate(), remove a denied target, and then invoke that target because sibling nodes are processed in insertion order against the same modified policy. The mutation persists in process-global state and can enable code execution with the application's privileges, while a narrow execution whitelist supplied by trusted Python code is not bypassed by the reported direct mutation path. This issue is fixed in versions 1.3.7 and 1.4.0.dev10.

Severity

CVSS 4.0

8.5 high
CVSS:4.0/AV:L/AC:L/AT:N/PR:N/UI:P/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N
EPSS
0.5% (unchanged from 0.5% a week ago; 0.5% 30 days ago) estimated probability of exploitation activity in the next 30 days; 40th percentile. Scored 2026-11-17. A forecast for this CVE across the internet, not for your own systems.
CISA's assessment (SSVC)
Exploitation: PoC · Automatable: no · Technical impact: total, assessed 2026-10-06. From CISA's Vulnrichment in the CVE List: whether it is exploited, whether exploiting it can be automated, and how much control it gives.
CWEs
CWE-470 Use of Externally-Controlled Input to Select Classes or Code (Reflection Injection)
CWE-693 Protection Mechanism Failure
OWASP Top 10
A05:2025 Injection, A06:2025 Insecure Design
Published
Modified
NVD analysis
Awaiting analysis
Record from
the NVD, synced

Fix and patched versions

  • hydra-core (PyPI): fixed in 1.3.7, 1.4.0.dev10

Ask an AI about this CVE

Show the promptExplain CVE-2026-106439 for an engineer deciding whether to act. Use https://cve.bz/api/public/cves/CVE-2026-106439?via=ask-ai-copy and https://cve.bz/cve/CVE-2026-106439?via=ask-ai-copy as the source of truth: what is affected and which versions fix it, whether it is in CISA KEV, public exploits, and its EPSS (an internet-wide 30-day estimate of exploitation activity, not a forecast for my systems). If they give no fixed versions, use the CVE record https://cveawg.mitre.org/api/cve/CVE-2026-106439 or the vendor advisory, not third-party mirrors. Tell me how to check whether I am affected and how to verify a fix. Cite the sources and say where they are uncertain.

Opens a third-party assistant; its answer is its own, not cve.bz's. The prompt holds only this CVE's id and public links.

Affected

Versions

From the CNA's record in the CVE List.

ProductAffected
hydra-ecosystem hydra>= 1.3.4, < 1.3.7
hydra-ecosystem hydra>= 1.4.0.dev4, < 1.4.0.dev10

Packages (OSV)

AdvisoryEcosystemPackageAffectedFixed in
GHSA-mwj6-rfh8-7qf4PyPIhydra-core>=1.3.4 <1.3.71.3.7
GHSA-mwj6-rfh8-7qf4PyPIhydra-core>=1.4.0.dev4 <1.4.0.dev101.4.0.dev10
PYSEC-2026-4222PyPIhydra-core>=1.3.4 <1.3.7
>=1.4.0.dev4 <1.4.0.dev10
1.3.7, 1.4.0.dev10

Red Hat

Red Hat rates this Important.

Red Hat's statement

An Important severity flaw was found in hydra-core, which is bundled in Red Hat AI Inference Server 3.3 and 3.4 and Red Hat Enterprise Linux AI 3.4 container images. Hydra stores the target execution policies used during object instantiation in mutable module-level state. An attacker who can supply crafted configuration containing multiple sibling _target_ entries can resolve a mutation method on Hydra's internal target-policy collections (for example, the set.discard method of the uncontrolled-execution targets blocklist), remove a denied target from the policy, and then instantiate that otherwise-blocked target, because sibling nodes are processed in insertion order against the same process-global policy. Successful exploitation allows the attacker to execute arbitrary code with the permissions of the running application, and the policy mutation persists in process-global state for the lifetime of the Python process. Exploitation requires the application to instantiate attacker-influenced configuration on the legacy path without an execution whitelist; a narrow execution whitelist supplied by trusted Python code is not bypassed by the reported direct mutation path. A fix for the affected Red Hat products is being pursued. Red Hat rates this flaw as Important (CVSS 3.1 score 7.8); the upstream CVSS v4 score of 8.5 differs due to CVSS version and user-interaction metric differences rather than a different view of the vulnerability.

Red Hat's mitigation

Avoid passing configuration from untrusted sources to Hydra's instantiate() or to Hydra-controlled Python logging configuration; restrict configuration input to trusted sources. On affected 1.4 development releases, applications can also supply a trusted, narrowly scoped execution whitelist from Python code, and the whitelist itself must not be derived from untrusted configuration. Restart any long-running process that may already have instantiated untrusted configuration, because a policy mutation persists in process-global state for the lifetime of the process. The upstream fix is available in hydra-core 1.3.7 on the stable line and 1.4.0.dev10 on the development line; for the affected Red Hat products, apply product updates containing the fixed version when they become available.

ProductPackageStateAdvisory
Red Hat AI Inference Server 3.5rhaii/vllm-cuda-rhel9:1791473631Fixed 2026-10-13RHSA-2026:82368
Red Hat AI Inference Server 3.5rhaii/vllm-cpu-rhel9:1791473614Fixed 2026-10-13RHSA-2026:82415
Red Hat Enterprise Linux AI 3.5rhelai3/disk-image-cuda-rhel9:1793206758Fixed 2026-10-29RHSA-2026:92449
Red Hat Enterprise Linux AI 3.5rhelai3/bootc-aws-cuda-rhel9:1793194817Fixed 2026-10-29RHSA-2026:92452
Red Hat Enterprise Linux AI 3.5rhelai3/bootc-azure-cuda-rhel9:1793194815Fixed 2026-10-29RHSA-2026:92452
Red Hat Enterprise Linux AI 3.5rhelai3/bootc-cuda-rhel9:1793135861Fixed 2026-10-29RHSA-2026:92452
Red Hat Enterprise Linux AI 3.5rhelai3/bootc-gcp-cuda-rhel9:1793194815Fixed 2026-10-29RHSA-2026:92452
Red Hat AI Inference Serverrhaiis/vllm-cpu-rhel9Not affected
Red Hat AI Inference Serverrhaiis/vllm-cuda-rhel9Not affected
Red Hat AI Inference Serverrhaiis/vllm-rocm-rhel9Not affected
Red Hat AI Inference Serverrhaiis/vllm-tpu-rhel9Not affected

Contains Red Hat Security Data © Red Hat, Inc., licensed under CC BY 4.0. Source: https://access.redhat.com/security/data. Modified: reformatted. Red Hat's page for CVE-2026-106439.

Attack patterns (CAPEC)

  • CAPEC-138 Reflection Injection Severity: Very High

The attack patterns MITRE's CAPEC relates to this record's weaknesses, as MITRE names them. Credits. More on the page of CWE-693 (17).

References

Check your stackJSON