backend-code-review
Pass
Audited by Gen Agent Trust Hub on Mar 26, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill uses clear, instructional language to define its scope and output format. It does not contain any attempts to bypass safety filters, override system instructions, or use jailbreak techniques.
- [DATA_EXFILTRATION]: No instructions or commands were found that access sensitive system files (e.g., credentials, SSH keys) or initiate unauthorized network requests to external domains.
- [COMMAND_EXECUTION]: The skill is entirely composed of Markdown documentation and does not include shell commands, script execution patterns, or instructions for the agent to run arbitrary code.
- [EXTERNAL_DOWNLOADS]: The skill does not reference or attempt to download any external dependencies, scripts, or remote content. All review rules are provided as local reference files.
- [SAFE]: The security rules mentioned in the 'General Review Rules' section are standard best practices for identifying vulnerabilities like SQL injection and SSRF during code audits.
Audit Metadata