CVE-2026-106440: CWE-470: Use of Externally-Controlled Input to Select Classes or Code ('Unsafe Reflection') in hydra-ecosystem hydra
Description
CVE-2026-106440 is a high-severity vulnerability in the hydra-ecosystem hydra framework affecting versions from 1.2.0 up to but not including 1.3.0 and certain 1.4.0 development releases. It involves unsafe reflection where externally controlled input can be used to select and execute arbitrary Python code via the hydra-optuna-sweeper package's custom_search_space configuration. This allows an attacker with control over Optuna sweep configuration or command-line overrides to execute code with the application's privileges, bypassing trusted execution whitelists. The issue is fixed in hydra versions 1.3.0 and 1.4.0.dev10.
CVSS v3.1
Score 7.8high
Affected software
hydra-ecosystem
hydra
Run on your own infrastructure? Check whether these packages are installed with threat-finder — our free open-source scanner.
Weaknesses
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
Hydra versions >=1.2.0 <1.3.0 and >=1.4.0.dev4 <1.4.0.dev10 contain a vulnerability in the hydra-optuna-sweeper package where the configuration-controlled dotted path in hydra.sweeper.custom_search_space is resolved using hydra.utils.get_method() without applying the usual execution policy. This allows an attacker who can control Optuna sweep configuration or command-line overrides to select importable Python code for execution with the application's privileges, effectively bypassing trusted execution whitelists. The vulnerability is categorized as CWE-470 (Use of Externally-Controlled Input to Select Classes or Code, 'Unsafe Reflection'). It is fixed in versions 1.3.0 and 1.4.0.dev10.
Potential Impact
An attacker with the ability to control Optuna sweep configuration or command-line overrides can execute arbitrary Python code within the context of the hydra application process. This can lead to full compromise of the application, including confidentiality, integrity, and availability impacts as indicated by the CVSS vector (C:H/I:H/A:H). The vulnerability bypasses trusted execution whitelists, increasing the risk of unauthorized code execution.
Mitigation Recommendations
This vulnerability is fixed in hydra versions 1.3.0 and 1.4.0.dev10. Users should upgrade to these versions or later to remediate the issue. No other mitigation or temporary workaround is indicated in the available data.
Technical Details
- Data Version
- 5.2
- Assigner Short Name
- GitHub_M
- Date Reserved
- 2026-10-06T16:49:40.590Z
- Cvss Version
- 3.1
- State
- PUBLISHED
Threat ID: 6ac546652cdf04f656d4a163
Added to database: 10/06/2026, 19:05:09 UTC
Last enriched: 10/06/2026, 19:18:21 UTC
Last updated: 10/06/2026, 19:19:23 UTC
Views: 4
Community Reviews
0 reviewsCrowdsource mitigation strategies, share intel context, and vote on the most helpful responses. Sign in to add your voice and help keep defenders ahead.
Want to contribute mitigation steps or threat intel context? Sign in or create an account to join the community discussion.
Actions
Updates to AI analysis require Pro Console access. Upgrade inside Console → Billing.
Need more coverage?
Upgrade to Pro Console for AI refresh and higher limits.
For incident response and remediation, OffSeq services can help resolve threats faster.
Latest Threats
Check if your credentials are on the dark web
Instant breach scanning across billions of leaked records. Free tier available.