Scan a prompt
Real embeddings · llama-3.3-70b-versatile · nomic-embed-text · cosine similarity
llama-3.3-70b · Groq nomic-embed-text
Scanned
0
Safe
0
Suspicious
0
Dangerous
0
Sample prompts
✅ Safe
💉 Prompt injection
🔓 Jailbreak
📤 Data exfiltration
🎭 Social engineering
Analyzing…
Analysis
Recommendation
Load a policy template
Or write your own security rules below
OWASP LLM Top 10 Corporate AI policy Healthcare / HIPAA
Embedding chunks…
Uses Groq nomic-embed-text-v1_5 — real 768-dim vectors
Indexed chunks
0 chunks · 0 embeddings
Build a knowledge base to see chunks here
Ask anything about your security policy or AI governance…
Searching knowledge base…
Suggested questions
What are the top prompt injection risks?
What should users never include in prompts?
How should we handle AI security incidents?
Penalties for policy violations?
Explain jailbreak attacks
Scan history
No scans yet — run your first analysis