skills/akillness/oh-my-gods/openhands/Gen Agent Trust Hub

openhands

Pass

Audited by Gen Agent Trust Hub on Aug 7, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the @openhands/agent-canvas package from the public NPM registry.
  • [COMMAND_EXECUTION]: The instructions involve executing the agent-canvas CLI tool to manage agent frontends and backends.
  • [REMOTE_CODE_EXECUTION]: The skill describes how to run the OpenHands agent, which is designed to execute code in a local or sandboxed environment.
  • [DATA_EXFILTRATION]: The skill includes a security notice warning the user that the agent has full filesystem access when run locally without a sandbox, encouraging the use of hardened configurations.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 7, 2026, 08:19 AM
Security Audit — agent-trust-hub — openhands