gcp-firestore-deep
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill uses standard, well-documented
gcloudCLI commands for Google Cloud Firestore administration and monitoring. All operations are scoped to the GCP environment and follow best practices for resource discovery. - [SAFE]: No evidence of hardcoded credentials, sensitive local file access (such as
~/.sshor.envfiles), or data exfiltration to untrusted domains was found. Network interactions are limited to official Google Cloud API endpoints. - [SAFE]: The skill includes comprehensive 'Anti-Hallucination Rules' and 'Counter-Rationalizations' that guide the AI agent toward safer and more reliable behavior by enforcing a 'discovery-first' workflow and preventing assumptions about resource names.
- [SAFE]: Shell operations, including the use of parallel background jobs and
jqfor JSON processing, are used appropriately for optimization and data transformation without introducing security risks like command injection.
Audit Metadata