cloud-pentest
Pass
Audited by Gen Agent Trust Hub on May 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill uses official cloud provider CLI tools (aws, az, gcloud) to query infrastructure metadata, including IAM policies, storage configurations, and network firewall rules. This behavior is consistent with the skill's primary function of cloud auditing.
- [EXTERNAL_DOWNLOADS]: The skill installs established third-party security auditing tools such as ScoutSuite, Prowler, and Pacu via the standard Python package manager (pip). These are reputable tools within the security community.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted metadata from cloud environment resources (such as resource names and policy descriptions) and incorporates this data into a final report.
- Ingestion points: Cloud API responses retrieved via CLI commands in SKILL.md and REFERENCE.md.
- Boundary markers: Minimal; the report uses headers but does not explicitly wrap external data with instruction-ignore delimiters.
- Capability inventory: Shell command execution (via subprocess) and file writing across all steps.
- Sanitization: The skill formats data using JSON tools and simple Python scripts but does not perform specific sanitization of the retrieved metadata before writing it to the report.
Audit Metadata