confidant
Warn
Audited by Snyk on Feb 19, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.80). The skill requires running "npx @aiconnect/confidant" at runtime, which fetches and executes code from the npm registry (e.g., https://registry.npmjs.org/@aiconnect/confidant) and is a required dependency for the secret-handoff flows.
Audit Metadata