Results for “system-prompt”

Wire Promptfoo and DeepTeam into CI/CD for automated regression red-teaming of LLM apps against OWASP LLM Top 10 and OWASP Agentic presets, failing the build when jailbreak or injection vulnerabilities regress.

Detect prompt injections and jailbreak attempts in LLM applications using Meta's 86M parameter classifier. Filter user inputs, third-party data, and RAG documents with low latency and multilingual support.

Detect and defend against prompt injection hidden in documents, web pages, and images consumed by an agent.

Automate multi-turn adversarial conversations against LLM agents using Microsoft PyRIT, including Crescendo and Tree-of-Attacks-with-Pruning (TAP) attack chains with scorer feedback loops.

OS command injection occurs when user input is passed unsanitized to a system shell via dangerous APIs: Java `Runtime.exec()`, Python `os.system/subprocess`, PHP `system/shell_exec/exec/proc_open`, C `system/exec`. Detect via pipe `|`, semicolon `;`, `&&`, `||`, backtick, `$()` operators, and time-delay payloads (`sleep 5`). Tools: Commix, Burp Suite, OWASP WebGoat.

Assess AI/ML systems for prompt injection, jailbreak vulnerabilities, model inversion risk, data poisoning exposure, and agent tool abuse, with MITRE ATLAS mapping and guardrail recommendations.

Performs interactive dynamic malware analysis using the ANY.RUN cloud sandbox to observe real-time execution behavior, interact with malware prompts, and capture process trees, network traffic, and system changes.