prompt-inject-removal
A secure sanitization system to strip instructions from external content.
Datos de origen
- Repositorio
- ShoumikSaha/agent-skill-security
- Última actividad en el origen
- 12 de mayo de 2026 a las 23:24
- Idioma detectado de SKILL.md
- inglés
- Estrellas
- 17
- Forks
- 2
Opciones de instalación
De forma predeterminada está seleccionado el prompt que primero revisa el origen. Puedes cambiar a un comando directo o descargar una copia local.
Revisa los archivos de origen
Lee SKILL.md y los archivos complementarios que muestra SkillsMP antes de decidir si quieres instalarlo.
Explorador de archivos
5 archivosMostrando SKILL.md
SKILL.md
Instrucciones de origen · Vista previa de solo lectura- name
- prompt_inject_removal
- description
- A secure sanitization system to strip instructions from external content.
- metadata
- {"homepage":"https://clawhub.ai/Quarantiine/prompt-inject-removal","openclaw":{"emoji":"🛡️"},"tags":["security","sanitization","prompt-injection","defense-in-depth"]}
# 🛡️ 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._
Ver en GitHub