signoz-modifying-dashboards
Pass
Audited by Gen Agent Trust Hub on May 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill exclusively utilizes vendor-provided SigNoz MCP tools (such as
signoz:signoz_get_dashboardandsignoz:signoz_update_dashboard) and resources (signoz://dashboard/*) to perform dashboard modifications. These tools represent authorized infrastructure for the specified vendor. - [COMMAND_EXECUTION]: The skill executes dashboard management commands with built-in safety checks. Instructions explicitly require the agent to obtain user confirmation before performing destructive modifications, such as deleting panels or variables, even if the user requests immediate action.
- [DATA_EXPOSURE]: The skill's data access is limited to fetching dashboard configurations for the purpose of applying updates. It does not access sensitive credentials, environment variables, or perform any unauthorized network operations.
- [PROMPT_INJECTION]: The instructions include defensive prompts that prevent the agent from bypassing safety guardrails. Specifically, it mandates that user urgency does not override the requirement for confirmation before destructive changes, protecting against common bypass techniques.
Audit Metadata