hebrew-chatbot-builder
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides logic for chatbots to process free-text Hebrew input from external users via WhatsApp and Telegram APIs. This ingestion of untrusted data creates a potential surface for indirect prompt injection. Ingestion points include handle_webhook in scripts/whatsapp-webhook-handler.py and handle_text in scripts/telegram-bot-scaffold.py which receive messages from external platforms. The skill documentation explicitly warns about prompt injection (referencing OWASP LLM01) and advises developers to wrap user text in clear delimiters. The capability inventory shows scripts performing network requests to official platform APIs (Meta Graph API and Telegram Bot API). The skill provides morphological analysis guidance but relies on the developer to implement strict LLM instruction guardrails.
- [COMMAND_EXECUTION]: The skill includes Python scripts intended for local execution to run a Flask-based webhook server and a Telegram bot client. These are contained in scripts/whatsapp-webhook-handler.py and scripts/telegram-bot-scaffold.py. The Flask implementation correctly disables the Werkzeug debugger and binds to 127.0.0.1 to prevent unauthorized remote access during development.
Audit Metadata