autoresearch

Warn

Audited by Snyk on Apr 19, 2026

Risk Level: MEDIUM
Full Analysis

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

  • Potentially malicious external URL detected (high risk: 0.90). The skill explicitly runs a remote install script with curl -LsSf https://astral.sh/uv/install.sh | sh (and also clones https://github.com/karpathy/autoresearch which provides code that is later executed), so external content is fetched-and-executed at runtime and is required for the workflow.

Issues (1)

W012
MEDIUM

Unverifiable external dependency detected (runtime URL that controls agent).

Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 19, 2026, 02:13 AM
Issues
1