monte-carlo-performance-diagnosis
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the Monte Carlo API, including job names, query text snippets, and error messages (found in
SKILL.mdworkflow steps). - Ingestion points: Data enters the agent's context through tools like
get_jobs_performance,get_top_slow_queries, andget_change_timeline. - Boundary markers: The instructions do not specify the use of delimiters or 'ignore' warnings for the data returned by these tools.
- Capability inventory: The skill is restricted to read-only tool calls and lacks high-risk capabilities like shell execution, file system writes, or direct network exfiltration.
- Sanitization: There is no evidence of sanitization or filtering applied to the external content before it is processed by the agent.
Audit Metadata