Security Reports

Live Security Telemetry

LIVE

Real events captured by the PromptShield security gateway.

Requests

0

Blocked

0

Review

0

Allowed

0

Avg Risk

0 / 100

LLM Forwarded

0

0 successful

Decision Distribution

Run a Firewall scan to populate telemetry.

AI Security Performance

Avg AI Confidence
0%

Gemini semantic analysis confidence across captured requests.

Gateway Performance

Average Latency
0 ms
LLM Requests0
Successful LLM Responses0

Detected Threat Distribution

No live threats recorded yet.

Security Event Stream

Latest firewall decisions captured locally.

0 EVENTS

No telemetry yet

Go to the AI Firewall and inspect a prompt. Every security decision will appear here automatically.

Telemetry is privacy-preserving: PromptShield stores the prompt length and SHA-256 fingerprint, not the raw prompt content. Current telemetry is stored locally in the browser for this demo.

Security Summary

Demo dataset

Prompts Analyzed

1,284

Threats Detected

347

Attacks Blocked

298

Critical Threats

26

Threat Breakdown

Direct Prompt Injection142 (41%)
Jailbreak98 (28%)
System Prompt Extraction54 (16%)
Instruction Override33 (9%)
Data Exfiltration20 (6%)
Obfuscation5 (1%)

Recent Threat Reports

Threat / TimestampSeverity

PromptShield telemetry is locally stored for this demonstration. Production deployments should use a server-side security event store.