gcloud
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill includes instructions to download and install the Google Cloud SDK using the official installation script:
curl https://sdk.cloud.google.com | bash. This targets a well-known service provided by Google. - [COMMAND_EXECUTION]: The skill facilitates extensive command-line operations via the
gcloud,bq, andgsutilutilities. These include resource creation, deletion, authentication management, and service configuration across Google Cloud Platform. - [INDIRECT_PROMPT_INJECTION]: The skill contains instructions for reading external data that could be attacker-controlled.
- Ingestion points: Data is ingested through commands like
gcloud logging read,gcloud run services logs read,gcloud storage cp, andgcloud secrets versions access(referenced inSKILL.mdandreferences/SERVICES.md). - Boundary markers: No explicit delimiters or boundary instructions are provided for handling the output of these commands.
- Capability inventory: The skill has the capability to execute shell commands, write to the file system (e.g.,
gcloud iam service-accounts keys create key.json), and perform network operations via the Google Cloud APIs. - Sanitization: There is no evidence of specific sanitization or filtering of the content retrieved from external Google Cloud services before it is presented to the agent context.
Audit Metadata