Stop prompt injection before it reaches your model

AiDren sits between your app and OpenAI or Anthropic — blocking injected prompts, scanning model files for malicious code, and watching your agent's outbound connections, before anything reaches production.

Start free trial

Your AI agent trusts too much

It calls untrusted APIs with your credentials. It loads model files pulled from public hubs, sight unseen. It makes its own outbound connections, unmonitored. Any one of those is a single point of failure — and until now, nothing was watching all three.

One security layer, three jobs

Prompt-injection blocking

A judge model screens every request before it reaches OpenAI or Anthropic. Injected instructions get blocked; clean traffic passes through untouched.

Malicious model-file scanning

Every pickle-format model file pulled through AiDren from Hugging Face Hub is statically scanned for dangerous opcodes before it ever reaches disk.

Egress monitoring

The worker-agent package watches your agent's own outbound connections and flags anything unexpected — without ever blocking a real one.

Live in five minutes

Add your OpenAI or Anthropic key

Paste in your real OpenAI or Anthropic API key. AiDren uses this to forward your requests on to the real provider — it's encrypted and never shown again.

Create a proxy key

This is the key your app uses instead of your real OpenAI/Anthropic key.

Point your app at AiDren

Change your base URL, keep everything else about the request the same:

curl https://api.aidren.co.uk/v1/chat/completions \
  -H "Authorization: Bearer YOUR_AIDREN_PROXY_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "gpt-4o", "messages": [{"role": "user", "content": "Hello"}]}'

14-day free trial — no card required

£19/mo after

Start free trial

Questions

What is prompt injection?

Prompt injection is when someone hides instructions inside content your AI reads — a webpage, an email, a file — hoping your model follows them instead of you. AiDren screens every request before it reaches your model, catching injected instructions before they can hijack your agent.

Will this slow my app down?

AiDren adds one fast classification pass before your request reaches OpenAI or Anthropic — typically well under the round-trip time to the model itself. Clean traffic passes straight through.

Which LLM providers are supported?

OpenAI and Anthropic today, with the same drop-in proxy pattern for both. More providers are coming.

Do you see my data?

Your requests pass through AiDren to make a block/allow decision and are never stored beyond the decision log you see in your own dashboard. Your API keys are encrypted at rest and never logged in plain text.

What happens after the trial?

Your 14-day trial needs no card up front. When it ends, upgrade to keep your proxy running — if you don't, AiDren pauses your traffic rather than silently letting it through unprotected.