Monitoring Dashboard Audit

Monitoring infrastructure assessment covering Grafana dashboard analysis, PromQL query validation, alert rule evaluation, SLA/SLO reporting review, and Prome...

This skill appears to be a legitimate read-only monitoring audit, but there are a few practical mismatches you should address before installing or running it: - Environment variables and least privilege: The CLI examples use GRAFANA_TOKEN, GRAFANA_URL, PROM_URL/AM_URL, yet the skill metadata declares none. Confirm with the author which env vars are required and supply a token scoped to Viewer (Grafana) and minimal Prometheus access. Never provide admin-level tokens if Viewer/Read-only is sufficient. - Local file access: The references include promtool commands targeting /etc/prometheus/*.yml. If you run the skill from an agent that has filesystem access, it could read local Prometheus configs. Ensure you only run the skill on hosts where such access is appropriate, or run the audit manually and share only the necessary artifacts. - Notification/contact data: The skill lists API calls that enumerate contact points and silences. These responses can include contact details and comments — treat outputs as sensitive and avoid sending them to external endpoints. - Autonomy and testing: Because this is an instruction-only skill, prefer to run it interactively the first time (not fully autonomous) and in a safe environment (staging read-only account) to verify the exact data it will access and the required env vars. Ask the publisher to update metadata to explicitly declare required env vars and any config paths the skill will read; that makes permission decisions explicit. If the author cannot provide clear required-env metadata or insists on needing broad credentials or system-wide file access without justification, treat the skill as higher-risk and avoid installing it in production environments.

Type: OpenClaw Skill Name: monitoring-dashboard-audit Version: 1.0.0 The monitoring-dashboard-audit skill bundle is a comprehensive tool for evaluating Grafana and Prometheus infrastructure. It provides structured procedures and reference commands (SKILL.md, cli-reference.md) for performing read-only assessments of dashboards, alert rules, and data source health. The logic is entirely consistent with its stated purpose, lacks any evidence of data exfiltration or malicious intent, and correctly identifies its safety tier as read-only.