agentbay-aio-skills
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is safe and implements code execution capabilities through a remote, isolated sandbox utilizing the AgentBay SDK.
- [COMMAND_EXECUTION]: The commands defined in the skill are restricted to setting up a local Python virtual environment (
.venv) and calling the dedicated execution script (scripts/run_code.py). No arbitrary or dangerous system commands are run on the host. - [CREDENTIALS_UNSAFE]: The handling of API keys follows best practices by either reading from the environment variable
AGENTBAY_API_KEYor a dedicated configuration file (~/.config/agentbay/api_key). It does not hardcode credentials or expose unrelated sensitive configuration files.
Audit Metadata