developer-device-platform-basics
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill makes use of standard command-line tools such as
gcloud,curl, andadb. These are used to authenticate with Google Cloud, enable necessary services, and interact with remote devices. All commands are transparent and directly related to the skill's primary function of device management. - [EXTERNAL_DOWNLOADS]: The skill specifies dependencies in a standard
requirements.txtfile, includinggoogle-cloud-devicestreaming,google-auth, andabsl-py. These are official, well-known libraries maintained by a trusted organization and are installed from standard package registries. - [DATA_EXFILTRATION]: While the skill manages an
ACCESS_TOKEN, it is generated from the user's localgcloudsession and used exclusively to authenticate requests togoogleapis.com. This is a standard and safe practice for cloud infrastructure management and does not constitute a data exfiltration risk. - [DYNAMIC_EXECUTION]: The skill includes a helper command that uses
python3 -cto parse and filter JSON output from acurlrequest. The Python code is a static, one-line script used for simple data processing of API responses, which is a common and safe technique in developer tools. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from Google Cloud API responses (device lists and session statuses). While this creates a data ingestion surface, the source is a trusted cloud provider, and the skill includes explicit user confirmation requirements for actions that incur billing or create resources, mitigating potential risks.
Audit Metadata