AGENT LAB: SKILLS

cloudflare-troubleshooting

Fail

Audited by Snyk on Feb 15, 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 explicitly asks the agent to collect the Cloudflare Global API Key/API Token and shows curl examples that embed X-Auth-Key/X-Auth-Email headers (including example secret values) so the agent would need to handle/output secrets verbatim.
Audit Metadata
Risk Level
HIGH
Analyzed
Feb 15, 2026, 08:27 PM