1
1
South Africa
1 year of experience
I'm ScoField, a 2nd-year BIT student, aspiring AI/ML engineer, and passionate innovator in agentic AI and automation. I build intelligent systems that blend real-time data with smart decision-making — from SaaS tools to AI-powered assistants. Currently leading and contributing to projects like ClariData and Guardrail Sentinel, I thrive at the intersection of code, strategy, and impact.

Guardrail Sentinel is a next-generation AI security tool designed to detect prompt injection vulnerabilities in AI-powered systems, including large language models (LLMs), chatbots, and agents. Whether you're running a public-facing AI assistant or deploying private LLMs via API, Guardrail Sentinel offers automated audits, guardrail enhancement suggestions, and ethical disclosures to help you stay secure, compliant, and competitive. Built with Trae Key Features: Endpoint Discovery & Classification Submit any LLM/chatbot API endpoint. We’ll analyze and classify its model type and surface metadata for risk evaluation. Prompt Injection Testing Agent We simulate attacks using advanced prompt templates (logic bombs, DAN-style bypasses, and recursive queries). Weakness Analysis & Guardrail Suggestions Get AI-driven insights, clarity on model behavior under attack, and proactive advice for strengthening your model’s defenses. Automated Reporting & Ethical Disclosure Generate shareable reports and use built-in templates for responsible disclosure to vendors or internal stakeholders.
15 Jun 2025