skills/mukul975/anthropic-cybersecurity-skills/implementing-cloud-workload-protection/Gen Agent Trust Hub
implementing-cloud-workload-protection
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses AWS Systems Manager (SSM) to execute shell commands on remote EC2 instances for monitoring purposes. This capability is central to the skill's functionality as seen in
scripts/agent.pyandSKILL.md. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from remote instances that is then stored in a report.
- Ingestion points:
scripts/agent.pyviassm.get_command_invocation. - Boundary markers: None present in the generated report or processing logic.
- Capability inventory: The skill can write to the local filesystem and interact with AWS APIs.
- Sanitization: No sanitization is performed on the output retrieved from remote commands.
Audit Metadata