Agent Firewall
Web Application Firewall (WAF) for AI agents. Rule-based traffic filtering with quarantine, rate limiting, and automatic blocking.
Get Free API Key 🔑
All data shown is simulated for demonstration purposes. Live data requires an API key.
ONLINE
Status
2.3ms
Latency
99.99%
Uptime
89
Blocked / 24h
Active Rules
| # | Rule | Action | Hits / 24h | Status |
|---|---|---|---|---|
| 1 | Identity verification required | Block if unverified | 342 | Active |
| 2 | Trust score minimum (50) | Quarantine below threshold | 12 | Active |
| 3 | Injection pattern detection | Block + alert | 89 | Active |
| 4 | Rate limiting (100 req/min) | Throttle | 67 | Active |
| 5 | Blocklist enforcement | Deny known threats | 23 | Active |
| 6 | Geo-restriction | Allow/deny by region | 0 | Config |
| 7 | Content policy enforcement | Filter prohibited content | 31 | Active |
Top Blocked Threats
Prompt Injection Cluster
47 blocked attempts from 3 unique agents targeting injection patterns.
47 blocked
Rate Limit Violations
23 agents throttled for exceeding 100 req/min threshold.
23 throttled
Deployment Modes
Inline Proxy
Route all traffic through the firewall. Full control, no code changes.
Sidecar
Deploy alongside your agent. Kubernetes-native.
API Gateway
Plug into Kong, Nginx, or AWS API Gateway.
SDK Middleware
Add as middleware in your existing agent code.
API Example
// Check if a request should be allowed
const decision = await ac.firewall.check({
agentDid: 'did:agent:incoming',
request: requestPayload
});
console.log(decision.action); // 'allow' | 'block' | 'quarantine'
console.log(decision.rule); // which rule triggered
// Add a custom rule
await ac.firewall.rules.create({
name: 'block-data-harvest',
condition: 'category == "data_harvest"',
action: 'block'
});
