monte-carlo-prevent

Pass

Audited by Gen Agent Trust Hub on Apr 14, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill contains an indirect prompt injection attack surface where data from an external API is incorporated into the agent's context.\n
  • Ingestion points: The agent ingests external data using the getQueryData, getAlerts, and getTable tools as described in references/workflows.md.\n
  • Boundary markers: The instructions do not specify the use of delimiters or 'ignore' warnings when processing data retrieved from the Monte Carlo API.\n
  • Capability inventory: The agent possesses the ability to write and edit local files and execute shell commands via the montecarlo CLI tool.\n
  • Sanitization: There is no description of sanitization, filtering, or validation processes for the data retrieved from external sources.\n- [COMMAND_EXECUTION]: The skill automates the execution of local command-line tools.\n
  • Evidence: Workflow 2 in references/workflows.md instructs the agent to run the command montecarlo monitors apply --auto-yes.\n
  • Notes: The inclusion of the --auto-yes flag allows the command to bypass manual confirmation steps, increasing the agent's autonomy during resource configuration.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 14, 2026, 02:10 PM