Detecting AI Model Prompt Injection Attacks

Detects prompt injection using regex signature matching, heuristic scoring for structural anomalies, and DeBERTa-based transformer classification, flagging direct injections (system-prompt overrides, role-play escapes) and indirect injections (encoded payloads, obfuscation) per OWASP LLM Top 10 (LLM01:2025). Use for input validation layers in chatbots/agents/RAG pipelines, or for retrospectively classifying injection attempts in logs or incident investigations.

26zl Updated 95 repo stars
Cybersec Toolkit by 26zl: the security toolkit AI agents can drive. One command installs 670+ tools, and an MCP server runs them in a disposable Kata VM under r

File contents

26zl/cybersec-toolkit commit 89f21c8235

Frequently asked questions

npx get-skill add 26zl/detecting-ai-model-prompt-injection-attacks-2