skills/mukul975/anthropic-cybersecurity-skills/implementing-network-access-control-with-cisco-ise/Gen Agent Trust Hub
implementing-network-access-control-with-cisco-ise
Warn
Audited by Gen Agent Trust Hub on Mar 16, 2026
Risk Level: MEDIUMCREDENTIALS_UNSAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
- [CREDENTIALS_UNSAFE]: The file SKILL.md contains hardcoded example credentials for RADIUS and TrustSec configuration (e.g., C0mpl3x$3cretKey!, CtsP@ss).
- [COMMAND_EXECUTION]: The script scripts/agent.py uses the subprocess module to execute the curl command with user-supplied arguments for auditing Cisco ISE APIs.
- [COMMAND_EXECUTION]: The auditing script explicitly disables SSL/TLS certificate validation by using the -k (insecure) flag in curl command calls, facilitating potential man-in-the-middle attacks.
- [EXTERNAL_DOWNLOADS]: The script performs network requests to remote Cisco ISE ERS and MnT API endpoints to fetch policy and session data.
- [DATA_EXFILTRATION]: The script transmits user-supplied credentials (username and password) to a remote server host using basic authentication over potentially unverified connections.
Audit Metadata