Most Used Tags
Published Resources
/guardx:schedule — Automated Scheduled Scanning
Automate scheduled scans with customizable frequency and webhook notifications.
/guardx:multimodal
Test vision-capable models for vulnerabilities to image-based prompt injection.
Adversarial Evaluator Agent
Adversarial Evaluator Agent identifies gaps in GuardX test suites to enhance security testing.
/guardx:report — Generate an HTML, SARIF, JUnit XML, or PDF report
Generate comprehensive reports in multiple formats from scan results.
/guardx:mcp-audit — MCP Configuration Audit
Audit your MCP configuration for vulnerabilities and security risks.
/guardx:supply-chain — LLM Supply Chain Security Scanner
Scan supply chain vulnerabilities in LLM projects to detect CVEs, secrets, and backdoors.
GuardX Security Scanner Agent
GuardX Security Scanner Agent identifies vulnerabilities in AI system prompts and provides actionable assessments.
/guardx:probes — Explore the attack probe catalogue
Explore the comprehensive catalogue of attack probes used by GuardX.
/guardx:harden — Adaptive Guardrails Generator
Generate adaptive guardrails to strengthen system prompts based on scan findings.
/guardx:diff
Compare two GuardX scans to identify changes, highlighting new regressions and resolved findings.
/guardx:rag-security — RAG Corpus Poisoning & Exfiltration Testing
Test RAG pipelines for vulnerabilities like injection and data exfiltration.
/guardx:mcp-security — MCP Protocol Security Testing
Test the security of MCP protocols to identify vulnerabilities and risks.
/guardx:interpret — Interpret and present scan results
Interpret and present scan results with detailed vulnerability analysis.
/guardx:tool-exfil
Test agentic systems for vulnerabilities in tool exfiltration and injection attacks.
/guardx:compliance — OWASP / NIST / MITRE ATLAS / EU AI Act / Agentic Top 10 Mapping
Map compliance findings to OWASP, NIST, MITRE, and EU AI Act standards.
/guardx:extended-probes — FlipAttack / PAP / Roleplay Scanner
Test AI systems for vulnerabilities using advanced attack techniques like FlipAttack and PAP.
/guardx:agent-escalation — Inter-Agent Privilege Escalation Testing
Test multi-agent systems for privilege escalation vulnerabilities.
/guardx:test-review — Adversarial Test Review
Streamline adversarial test reviews with automated evaluations and feedback.
Test Writer Agent
Automate the creation of comprehensive Vitest unit test suites for GuardX modules.
GuardX
GuardX is a comprehensive LLM security scanner that identifies prompt injection and extraction risks before deployment.
/guardx:scan — Run a security scan on a system prompt
Run a security scan on system prompts to identify vulnerabilities quickly.
/guardx:red-team — Agentic Red Team Attack
Conduct adversarial red team tests on system prompts to identify vulnerabilities.
/guardx:test — GuardX Testing Assistant
Ad-hoc testing assistant for GuardX that supports functional and non-functional testing tasks.
/guardx:canary — Canary Token Management
Manage canary tokens to detect prompt leakage effectively.
/guardx:scan-endpoint — HTTP Endpoint Security Scan
Perform security scans on HTTP endpoints to identify vulnerabilities.
/guardx:handover — Generate a complete session handover note
Generate a comprehensive session handover note with essential details and insights.
/guardx:promptware — Promptware Kill Chain Simulator
Simulate prompt injection kill chains to evaluate adversarial control flows.
/guardx:history — Browse scan history
Quickly browse and retrieve your scan history with detailed results.