pippit-one-stop-skill

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

• Third-party content exposure detected (high risk: 0.90). This skill directly calls and ingests responses from the public Pippit API (https://www.pippit.ai) — see SKILL.md and scripts/nest/get_thread.py (and related submit_run.py/download_results.py) — and parses assistant messages, download URLs, interactions, and tool-call requests which the agent uses to drive downloads, uploads, scheduling, and other follow-up actions, so untrusted third-party content can materially influence behavior.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill calls the external Pippit API endpoints (default PIPPIT_BASE = https://www.pippit.ai and PIPPIT_HOME_URL = https://www.pippit.ai/home?) at runtime to submit user messages and receive generated assistant messages/clarification that directly drive prompts, interactions, and downloads, and the scripts require that service to operate.