agent-webex

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.90). The skill explicitly ingests user-generated Webex content via commands like agent-webex message list/message get (documented in SKILL.md and references/common-patterns.md) and the provided templates (e.g., templates/monitor-space.sh and pattern 28) show the agent reading message text and taking actions (auto-responding), so untrusted third-party messages can influence tool use and decisions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs running the npm package at runtime (e.g., "npx -y agent-messenger ..." / "pnpm dlx agent-messenger" / "bunx agent-messenger"), which will fetch and execute code from the npm registry (https://registry.npmjs.org/) during runtime, so remote code is downloaded and run as a required dependency.