together-sandboxes

Warn

Audited by Snyk on Mar 31, 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 calls the Together Sandboxes API (base https://api.together.ai and endpoint https://api.together.ai/tci/execute) at runtime to execute arbitrary Python code in remote sandboxes, so this external URL directly executes code and is a required dependency.

Issues (1)

W012
MEDIUM

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

Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 31, 2026, 07:24 PM
Issues
1