AInforce
Runtime AI Firewall

Every request through the gateway
is inspected inline.

An inline security layer between your application and the foundation model. Blocks prompt injections, jailbreaks, credential leaks, and malicious tool calls — before they reach the model or leave in the response.

Inline
Runs in-process, no added RTT
6
Threat categories
OWASP + MITRE
Framework-mapped
Design partner program
Your App
Runtime Guard
Foundation Model
inspect
94.2%
Allowed
3.8%
Flagged
2.0%
Blocked
inline inspection — no added round-trip to your model
Architecture

One line between your app and the model

Your application

Any LLM-powered product — chatbot, agent, copilot, RAG pipeline. No changes to your architecture.

Runtime Guard

Inspects every input and output inline. Classifies threats, scores risk, and enforces your policy with no added round-trip to the model.

Foundation model

GPT-4o, Claude, Gemini, Llama — any model, any provider. The guard is model-agnostic.

Live demo

Test the guard against real attack patterns

Select an example or write your own. The check runs against the real detection engine.

Context:
Verdict

Run a check to see the verdict

Detection coverage

Six threat categories, mapped to OWASP and MITRE ATLAS

critical

Prompt injection

Attempts to override system instructions via user input — including direct instruction overrides, authority spoofing, and role hijacking.

MITRE AML.T0051OWASP LLM01
high

Jailbreak

Persona manipulation, DAN mode, developer mode unlocks, and hypothetical framing designed to remove model safety constraints.

MITRE AML.T0054OWASP LLM01
critical

PII exfiltration

Credit card numbers, SSNs, email addresses, and health identifiers detected in transit — in both input and model output.

MITRE AML.T0025OWASP LLM06
critical

Credential exposure

API keys, bearer tokens, database connection strings, AWS keys, GitHub tokens, and Stripe secrets flagged before they reach the model.

MITRE AML.T0025OWASP LLM06
critical

Malicious tool calls

Shell injection, path traversal, SSRF to metadata services and private IP ranges, destructive SQL, and code execution in tool arguments.

MITRE AML.T0052OWASP LLM07
high

Indirect injection

Adversarial instructions embedded in documents, emails, or web pages retrieved by RAG pipelines — targeting your agent via external content.

MITRE AML.T0051.001OWASP LLM01
Threat feed

What the guard sees in production

Live
VerdictThreatApplicationRiskWhen
blockPrompt Injectionsupport-bot
95
2s
allowNo threatcode-assistant
2
4s
flagPII Exfiltrationhr-copilot
42
9s
blockJailbreaksales-agent
78
15s
allowNo threatdocs-bot
0
18s
blockMalicious Tool Callfinance-agent
99
23s
flagCredential Leakdevops-helper
55
31s
allowNo threatsupport-bot
4
35s
blockIndirect Injectionrag-agent
88
41s
allowNo threatcode-assistant
1
47s
Integration

Wraps any LLM call in two lines

The SDK works as a drop-in wrapper. Point it at any model provider — OpenAI, Anthropic, Google, Mistral, or a self-hosted endpoint. No changes to your prompt logic, no new infrastructure.

Inline inspection — no added round-trip to the model

Inspect inputs, outputs, and individual tool call arguments independently

Policy-as-code: configure block/flag/allow thresholds per application

Full audit log of every check — verdict, threats, risk score, latency

TypeScript / Node.js
import { RuntimeGuard } from "@ainforce/sdk";

const guard = new RuntimeGuard({ apiKey: process.env.AR_API_KEY });

// Drop-in wrapper — your existing LLM call unchanged
async function chat(userMessage: string) {
  // Inspect the input before it reaches the model
  const check = await guard.check({
    input: userMessage,
    context: "user",
  });

  if (check.verdict === "block") {
    return { error: "Request blocked by AI security policy." };
  }

  // Your existing LLM call
  const response = await openai.chat.completions.create({
    model: "gpt-4o",
    messages: [{ role: "user", content: userMessage }],
  });

  // Optionally inspect the output before returning it
  const outputCheck = await guard.check({
    input: response.choices[0].message.content ?? "",
    context: "assistant",
  });

  if (outputCheck.verdict === "block") {
    return { error: "Model output blocked — possible data exfiltration." };
  }

  return response;
}

Request a demo

See Runtime Security in your environment

We'll walk you through a live deployment and show you exactly what inline prompt inspection catches in your stack.

No purchase commitment. We use these details to respond to your request. Privacy policy