setup

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill asks the agent to accept or prompt for the user's API key and then embed that exact key into shell rc updates or commands (e.g., exporting LARKCI_API_KEY via echo or file edits), which forces the LLM to handle and output the secret verbatim.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs running "npm install -g @getlark/cli", which fetches and installs remote executable code from the npm registry (e.g. https://registry.npmjs.org/@getlark/cli) and then runs the getlark binary, so the runtime pulls and executes external code that the skill depends on.