setup-agent-tail

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill explicitly instructs the agent to preserve and wrap existing package.json run commands verbatim (embedding the original command inside the new agent-tail runner), which would force the LLM to read and output any secrets that were in those commands (API keys, tokens, or passwords) — a direct exfiltration risk.