protect-mcp-setup

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The hook commands pass the full "$TOOL_INPUT" and "$TOOL_OUTPUT" as literal command-line arguments to the CLI, which means any API keys, bearer tokens, or passwords contained in tool inputs/outputs would need to be included verbatim and thus exposed in commands/process arguments and logs.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill invokes npx protect-mcp@latest and npx @veritasacta/verify at runtime—which fetch and execute packages from the npm registry (e.g. https://www.npmjs.com/package/protect-mcp and https://www.npmjs.com/package/@veritasacta/verify)—so external code is fetched/executed at runtime and is a required dependency.