skills/mukul975/anthropic-cybersecurity-skills/implementing-network-access-control-with-cisco-ise/Snyk
implementing-network-access-control-with-cisco-ise
Fail
Audited by Snyk on Mar 19, 2026
Risk Level: HIGH
Full Analysis
HIGH W007: Insecure credential handling detected in skill instructions.
- Insecure credential handling detected (high risk: 1.00). The prompt contains literal secrets (e.g., RADIUS shared secret C0mpl3x$3cretKey!, CTS password CtsP@ss, test passwords) inside configuration snippets, so generating or reproducing those configs would require the LLM to handle/output secret values verbatim.
HIGH W008: Secret detected in skill content (API keys, tokens, passwords).
- Secret detected (high risk: 1.00). I scanned for literal, directly embedded credentials that look like real, usable secrets (not placeholders or obvious examples).
Flagged:
- "C0mpl3x$3cretKey!" — This string is repeatedly used as a RADIUS Shared Secret / server-key in device and switch configuration snippets (RADIUS Shared Secret, key 0 C0mpl3x$3cretKey!, client … server-key C0mpl3x$3cretKey!). It is a specific, non-placeholder value with mixed-case letters, numbers, and symbols and is used in authentication/RADIUS contexts, so it appears to be a real hardcoded secret.
Ignored (and why):
- "CtsP@ss" — used as cts credentials id … password CtsP@ss. This is short and resembles a low-security setup password/sample; per the guidance to ignore low-security/setup passwords, I treated this as a non-actionable example.
- "testpass" in the test aaa command — obvious test/example password; low-entropy placeholder — ignored.
- Other items (IP addresses, AD group names, usernames like radius-test, SNMP settings described generically, references to providing domain admin credentials) do not contain direct secret values or are documentation placeholders, so they were not flagged.
Conclusion: There is at least one real, embedded high-entropy secret present (C0mpl3x$3cretKey!).
Issues (2)
W007
HIGHInsecure credential handling detected in skill instructions.
W008
HIGHSecret detected in skill content (API keys, tokens, passwords).
Audit Metadata