plan-the-rollback
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill does not contain any malicious patterns or security risks. It functions as a structured template and set of guidelines for humans or AI agents to document technical rollback procedures.
- [NO_CODE]: There are no scripts (Python, Node.js, Shell) or automated commands included in the skill. The code snippets provided (e.g., SQL ALTER statements) are examples for documentation purposes rather than instructions for direct execution.
- [PROMPT_INJECTION]: No instructions were found that attempt to override the AI agent's safety protocols, role-play restricted personas, or bypass core instructions.
- [DATA_EXFILTRATION]: There is no evidence of commands that access sensitive system files (like SSH keys or environment variables) or attempt to transmit data to external servers.
- [OBFUSCATION]: The content is written in clear, plain text. No hidden characters, Base64 encoding, or homoglyph-based URL spoofing were detected.
- [REMOTE_CODE_EXECUTION]: The skill does not download or execute external dependencies, binaries, or remote scripts.
Audit Metadata