Making a safe request
Send a normal prompt through Vaikora. The request is evaluated against your policies and forwarded to your LLM provider, and the response is returned as usual.
Blocking a threat
Send a prompt that contains an injection or a leaked secret. Vaikora blocks it before it reaches the model and returns a decision explaining why.
Using SDKs
The Node.js and Python SDKs wrap the REST API so you can integrate with a few lines of code while keeping full streaming support.
Understanding the response
Every Vaikora response carries the decision, the matched policy, and a receipt so you can log, audit, and act on what happened.