tune-monitor
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from monitor reports and incident summaries (
get_monitor_report) to generate tuning recommendations. These recommendations can then be applied to modify production monitoring configurations. - Ingestion points: Monitor incident reports, troubleshooting summaries, and configuration details fetched from the Monte Carlo API (Phase 1).
- Boundary markers: The skill does not use explicit boundary markers for raw input data but generates its own structured analysis report (Phase 4).
- Capability inventory: The skill can create or update metric, SQL, table, and validation monitors, including modifying SQL queries and thresholds.
- Sanitization: The skill validates the monitor UUID input format (Phase 0) and validates generated YAML configurations against the official MaC JSON schema before application (Phase 5).
- [EXTERNAL_DOWNLOADS]: The skill downloads a JSON schema from an official vendor domain (
https://clidocs.getmontecarlo.com/mac/schema.json) to validate configuration changes. This is documented as a safe operation for verification purposes.
Audit Metadata