The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources and its own historical snapshots, creating a theoretical injection surface.
Ingestion points: Metric data and event names fetched via tool_run_report and tool_run_realtime_report, as well as previous snapshots retrieved through read_context_node.
Boundary markers: None explicitly defined for the interpolation of GA4 metrics into the output summaries.
Capability inventory: Accesses GA4 reporting tools, reads and writes to internal context nodes, and logs issues via create_finding.
Sanitization: No specific sanitization or validation logic is applied to the ingested metric strings or event names before reporting.

ga4-monitor