pp-ht-ml

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). This skill includes commands that reveal or export per-site "update_key"/write keys (e.g., keys export / "Reveal a single write key"), notes that secrets are stored locally (credentials.toml), and supports delivering command output to external webhooks — meaning an agent invoking those commands could receive and be asked to emit secret values verbatim (exfiltration risk).

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill requires installing a remote CLI and explicitly instructs running commands that fetch and execute remote code (e.g., "npx -y @mvanhorn/printing-press-library install ht-ml --cli-only" and "go install github.com/mvanhorn/printing-press-library/library/cloud/ht-ml/cmd/ht-ml-pp-cli@latest"), so these external package sources would execute code at install-time and are required for the skill to run.