portpilot-assistant

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

• Third-party content exposure detected (high risk: 0.70). The skill's workflow and scripts explicitly include a fallback to run "npx -y portpilot-cli@1" (see SKILL.md "Fallback to npx" and scripts/run_portpilot.sh), which would fetch and execute a package from the public npm registry so untrusted third‑party code/output could influence CLI results and subsequent agent actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's runtime script (scripts/run_portpilot.sh) falls back to invoking "npx -y portpilot-cli@1", which will fetch and execute remote package code from the npm registry (e.g. https://registry.npmjs.org/) at runtime if the bundled CLI is missing, so this is a remote-code execution dependency used during runtime.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 1.00). The skill explicitly requests one-time escalated/sandbox permissions to run its bundled CLI (and to retry read actions with escalation on permission errors) and includes write actions (free/init/config migrate) that modify system state, so it pushes the agent to obtain elevated privileges and perform state-changing operations.