MediumVulnerability

CVE-2026-4963: huggingface smolagents code injection in local Python executor evaluation

Source: NVD(opens in a new tab)
Identifier: CVE-2026-4963
Published: March 27, 2026
Record updated: October 9, 2026

Affected: smolagents <= 1.25.0.dev0
Fixed in: No fixed version was stated when the source was last read.
Known exploitation: Not listed in the CISA Known Exploited Vulnerabilities catalog at the last check.
EPSS: 0.6%

Summary

A weakness in huggingface smolagents 1.25.0.dev0, in the functions evaluate_augassign, evaluate_call and evaluate_with in src/smolagents/local_python_executor.py, allows code injection. It is an incomplete fix for CVE-2025-9959 and can be triggered remotely. A public exploit exists, and the vendor did not respond to early disclosure.

Mitigation

The source does not state a fix yet. Check the original advisory for updates.

Details

Scores and exploitation

CVSS score: 6.3(medium)
CVSS vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:L/I:L/A:L
Attack vector: network
Attack complexity: low
Privileges required: none
User interaction: required
EPSS: 0.6%Estimated probability that this CVE is exploited in the next 30 days, from the FIRST Exploit Prediction Scoring System. Read on November 9, 2026.
Disclosure date: March 27, 2026

Weaknesses and techniques

Weakness (CWE): CWE-74
CWE-94
Attack pattern (CAPEC): CAPEC-242

Affected

Packages: smolagents@<= 1.25.0.dev0
Vendors: HuggingFace

Classification

Attack type: Other
Attack sophistication: Moderate
AI component targeted: Agent
Impact: Confidentiality,
Integrity
Labels: security

How this record was produced

Summaries, mitigations and labels on this page are generated automatically from the source. The linked source is authoritative. The source gives this advisory a description and no title, so the title above is a headline generated from the source text.

Original source: https://nvd.nist.gov/vuln/detail/CVE-2026-4963

First tracked: March 27, 2026 at 02:07 PM ET, from NVD/CVE Database

Classifier: claude-haiku-5-5, prompt v4

Classifier confidence: 90% (self-assessed)

Affected packages in the Exposure Registry

Matched by package name and ecosystem. Each entry shows whether the package delegates to a language model and how many tracked packages depend on it.

smolagentsPyPILLM dependency since 2024-12-27 · 7 tracked dependents

Summary

Mitigation

Affected packages in the Exposure Registry

Related items