monitoring-grafana
Pass
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The agent is instructed to run a bundled discovery script (
discover.ts) using the Bun runtime to identify datasources, dashboards, and alerts in the Grafana environment. - [PROMPT_INJECTION]: The skill processes telemetry data from external sources, which presents a surface for indirect prompt injection.
- Ingestion points: Telemetry data, including Loki logs, Prometheus label names/values, and Grafana alert titles, are ingested into the agent context via tool outputs (described in
SKILL.md). - Boundary markers: Absent; the skill instructions do not define delimiters or specific system prompts to isolate external monitoring data from agent instructions.
- Capability inventory: Includes the execution of local scripts via
bun runand performing authenticated network operations to the Grafana API (SKILL.md,discover.ts). - Sanitization: Absent; the skill does not perform content sanitization or filtering of the ingested logs or metrics before they are interpreted by the agent.
Audit Metadata