prompt-inject-removal-p

A robust security layer that detects and removes prompt injection attempts from external content, ensuring your AI interactions remain safe and controlled. I...

🛡️ Prompt Inject Removal

This skill provides a secure way to summarize untrusted external content (web pages, articles, blogs) by routing it through a "Zero-Trust" sanitization prompt.

🚀 Setup & Configuration

This skill is powered by a local, hardened system prompt. No external API keys or complex configuration are required.

📐 Workflow (Sanitization)

1. Fetch: Raw content is retrieved via `web_fetch` or `browser`.
2. Delimit: The content is wrapped in `<untrusted_input_data>` tags.
3. Sanitize: The Main Agent processes the content using the rules in [PROMPT.md].
4. Ingest: Only the resulting sanitized summary is used in the conversation.

📖 Security Reference

• Detailed Security Docs: [references/security.md]
• Hardened System Prompt: [PROMPT.md]

---

Disclaimer: This is a defense-in-depth tool. While it significantly mitigates prompt injection risks, no prompt-based sanitization is 100% foolproof. Review sanitized data before performing state-changing actions.