backup-restore-drills
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and analyze external repository contents, such as backup tool documentation, data models, and runbooks. While this creates a surface for indirect prompt injection from untrusted files, the skill provides strong defensive guidance, emphasizing environment isolation and independent verification of state.
- [DATA_EXFILTRATION]: The instructions contain explicit prohibitions against unauthorized data access and exfiltration, such as 'Do not download real customer backups' and requirements to 'Disable production notifications, webhooks and outbound effects' during restoration drills.
- [SAFE]: The skill is entirely instructional and does not contain any executable scripts, shell commands, or network-enabled code. Its content promotes a 'least privilege' approach to disaster recovery testing, consistently requiring explicit approval for destructive actions or access to live environments.
Audit Metadata