linkedin-storytelling
Warn
Audited by Socket on Jul 22, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS: the writing functionality is coherent, and ColdIQ domain/API usage matches the stated purpose, but the skill also instructs the agent/user to run an external MCP package with an API key via mutable npx @latest. Because the exact @coldiq/mcp package provenance was not verified, the main risk is supply-chain plus credential forwarding rather than clear malicious intent.
Confidence: 84%Severity: 67%
Audit Metadata