Security Audit
gcloud
github.com/Mrc220/agent_flywheel_clawdbot_skills_and_integrationsTrust Assessment
gcloud received a trust score of 86/100, placing it in the Mostly Trusted category. This skill has passed most security checks with only minor considerations noted.
SkillShield's automated analysis identified 1 finding: 0 critical, 1 high, 0 medium, and 0 low severity. Key findings include Broad GCP CLI Access Enables Excessive Permissions.
The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. All layers scored 70 or above, reflecting consistent security practices.
Last analyzed on February 11, 2026 (commit c7bd8e0f). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.
Layer Breakdown
Behavioral Risk Signals
Security Findings1
| Severity | Finding | Layer | Location | |
|---|---|---|---|---|
| HIGH | Broad GCP CLI Access Enables Excessive Permissions The 'gcloud' skill provides comprehensive access to the Google Cloud CLI, covering a vast array of services including Compute Engine, Cloud Run, GKE, Cloud Functions, Cloud Storage, BigQuery, Pub/Sub, Secret Manager, IAM, Cloud Build, Artifact Registry, Logging, and App Engine. While the skill itself is documentation, its purpose is to enable an AI agent to execute these commands. If the underlying execution environment for the agent is configured with broad GCP permissions (e.g., a service account with 'Owner' or 'Editor' roles), the agent could perform highly destructive actions, exfiltrate sensitive data, or create/modify critical infrastructure without fine-grained control. This skill, by its nature, encourages or facilitates an agent operating with excessive permissions if not properly constrained. Implement strict access controls for the AI agent's execution environment. The service account used by the agent should adhere to the principle of least privilege, granting only the specific GCP permissions required for its intended tasks. Consider creating multiple, specialized 'gcloud' skills, each with a limited scope of commands, rather than a single, all-encompassing skill. Additionally, implement an approval mechanism or human-in-the-loop for sensitive `gcloud` operations. | LLM | SKILL.md:1 |
Scan History
Embed Code
[](https://skillshield.io/report/4ecdf9a021e78226)
Powered by SkillShield