skills/mukul975/anthropic-cybersecurity-skills/hunting-credential-stuffing-attacks/Gen Agent Trust Hub
hunting-credential-stuffing-attacks
Pass
Audited by Gen Agent Trust Hub on Mar 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements legitimate cybersecurity monitoring logic for detecting distributed authentication attacks. It analyzes login velocity, IP diversity, and user-agent patterns using pandas.
- [SAFE]: No network operations, remote code execution, or data exfiltration patterns were detected. The script only reads local log files and writes a JSON report.
- [SAFE]: Instructions and metadata do not contain any prompt injection attempts or obfuscated content.
- [SAFE]: External references in the documentation point exclusively to well-known and trusted sources, including OWASP and official Splunk documentation.
- [SAFE]: The skill relies on standard, verifiable Python packages (pandas, numpy) for data processing.
Audit Metadata