skills/mukul975/anthropic-cybersecurity-skills/detecting-cloud-threats-with-guardduty/Gen Agent Trust Hub
detecting-cloud-threats-with-guardduty
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The
scripts/agent.pyscript utilizessubprocess.runto execute AWS CLI commands for GuardDuty operations. This is a standard method for programmatic cloud management.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes findings from AWS GuardDuty which may contain externally controlled data (e.g., finding descriptions).\n - Ingestion points: Findings are retrieved via
list-findingsandget-findingsinscripts/agent.py.\n - Boundary markers: None present in the data processing logic to separate external finding content from agent instructions.\n
- Capability inventory: The agent can list, view, and archive findings through the AWS CLI.\n
- Sanitization: Basic JSON parsing is performed on findings, but no domain-specific sanitization or escaping is applied to finding text fields before reporting.
Audit Metadata