cxas-configurable-dashboards

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the cxas CLI via uv run to perform operations such as pull-dashboards, diff-dashboards, and push-dashboards. These commands interact with Google Cloud Platform APIs to manage dashboard resources, aligning with the skill's primary purpose.
  • [INDIRECT_PROMPT_INJECTION]: The skill acts as an interface for authoring dashboards, ingesting user-defined metrics and configuration requirements into dashboards.yaml. This constitutes a surface for indirect prompt injection where inputs influence the generated SQL queries and visualization specifications. * Ingestion points: Workspace configuration files (dashboards.yaml) and natural language requirements from the user. * Boundary markers: None present in the instruction set. * Capability inventory: Local file system writes and shell command execution (uv run cxas) to sync configurations with remote cloud services. * Sanitization: No explicit content sanitization or validation of SQL/Vega-Lite specifications is performed within the skill instructions or scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 02:19 PM
Security Audit — agent-trust-hub — cxas-configurable-dashboards