UTC
Critical

PickleScan's pkgutil.resolve_name has a universal blocklist bypass

CVE-2026-3490 · GHSA-vvpj-8cmc-gx39 · published 2026-03-03 · updated 2026-06-18

Severity
Critical
CVSS score
10.0
Chance of exploitation (EPSS)
0.82%
Fixed in
1.0.4
  • Exploitation Public proof of concept
  • Automatable Yes
  • Technical impact Total
  • CISA SSVC
What it is

This is a complete bypass of picklescan's security model. The entire blocklist — every module and function entry in unsafeglobals — is rendered ineffective. An attacker needs only use pkgutil.resolve_name as an indirection layer to call any Python function. This affects: HuggingFace Hub (uses picklescan) Any ML pipeline using picklescan for safety validation Any system relying on picklescan's blocklist to prevent malicious pickle execution

What to do

Upgrade picklescan to 1.0.4 or later.

1. Immediate: Add pkgutil to unsafeglobals: 2. Also block similar resolution functions: 3. Architectural: The blocklist approach cannot defend against indirect resolution gadgets. Even blocking pkgutil, an attacker could find other stdlib functions that resolve module attributes. Consider: Analyzing REDUCE arguments for suspicious strings (e.g., patterns matching "module:function") Treating unknown globals as dangerous by default Switching to an allowlist model

Affected
PackageRegistryAffectedFixed in
picklescanPyPI< 1.0.41.0.4

The impact, patch and workaround text comes from the advisory, shortened (GitHub Advisory Database, CC BY 4.0). Proof-of-concept and reproduction sections and code are left out; the original advisory has them.

Sources: GitHub Advisory Database, reviewed and malware advisories (CC BY 4.0, https://github.com/advisories); advisories published by Cursor (https://github.com/cursor/cursor/security/advisories); advisories published by OpenClaw (https://github.com/openclaw/openclaw/security/advisories); advisories published by vLLM (https://github.com/vllm-project/vllm/security/advisories); advisories published by Open WebUI (https://github.com/open-webui/open-webui/security/advisories); advisories published by n8n (https://github.com/n8n-io/n8n/security/advisories); advisories published by Flowise (https://github.com/FlowiseAI/Flowise/security/advisories); advisories published by Langflow (https://github.com/langflow-ai/langflow/security/advisories); advisories published by the MCP Python SDK project (https://github.com/modelcontextprotocol/python-sdk/security/advisories); advisories published by MLflow (https://github.com/mlflow/mlflow/security/advisories); CISA Known Exploited Vulnerabilities catalog (CC0). Data as of 2026-11-22. ShieldMap describes how each problem works and how to fix it; it does not publish exploit code. Report a correction to [email protected].