monte-carlo-asset-health
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes and displays external data fetched from Monte Carlo MCP tool outputs, such as alert types, status descriptions, and table metadata.
- Ingestion points: Untrusted data enters the agent context through the outputs of the
search,get_alerts, andget_tabletools as specified inreferences/workflows.md. - Boundary markers: The skill implements instruction-level boundaries in
SKILL.md, explicitly directing the agent to "Only report data returned by the tools" and "Do NOT call additional tools, do NOT infer or fabricate metrics." - Capability inventory: The skill is configured for data retrieval and reporting; it does not possess capabilities for arbitrary file system modifications, shell command execution, or unauthorized network operations.
- Sanitization: The skill mandates the use of specific output templates and structured formatting, which mitigates the risk of the agent misinterpreting ingested data as new instructions.
Audit Metadata