port-whisperer

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

• Third-party content exposure detected (high risk: 0.70). The skill's scripts/ports.sh explicitly falls back to running "npx --yes port-whisperer" (downloading/executing a package from the public npm registry) and the agent reads and acts on that tool's output to decide inspections/kills/clean operations, so untrusted third‑party package output could indirectly inject instructions influencing behavior.

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

• Potentially malicious external URL detected (high risk: 0.90). The script explicitly falls back to running "npx --yes port-whisperer" at runtime, which fetches and executes remote code from the npm registry (e.g., https://registry.npmjs.org/) so external content is executed and can control the skill's behavior.