agent-sandbox

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for E2B sandbox management, which provides secure, isolated execution environments by design.
  • [COMMAND_EXECUTION]: The skill uses tools like mcp__flow-nexus__sandbox_execute to run code. This functionality is the primary purpose of the skill and is confined to the isolated sandbox environment.
  • [CREDENTIALS_SAFE]: Example configurations for environment variables use generic placeholders (e.g., API_KEY: "key") rather than actual secrets.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 12:23 PM
Security Audit — agent-trust-hub — agent-sandbox