Products
Where ArmorIQ enforces policy on your agents, and how to pick the one that fits how they run.
Each product enforces the same policies you write in the Platform, at a different point in your agent's path.
ProductIntent Gateway
Policy on every tool call your agent's model returns. Change one base URL, no SDK.
ProductIntent ContainerHolds each agent task to its objective on files, processes and network calls, down to the kernel.
Connect an agent in one line
Point your agent's SDK at the gateway. Your provider key, prompts and tools stay as they are.
OPENAI_BASE_URL=https://gw.armoriq.ai/v1/t/<gateway-key>/a/<agent-id>Intent Gateway capabilities
Quickstart
Create a gateway key and send your first governed request.
How enforcement worksWhat the gateway does with each tool call, and what happens when ArmorIQ can't decide.
Monitor modeSee what policy would allow, hold and block on real traffic before anything is stopped.
Which one to start with
- Your agent calls a model API through an OpenAI, Anthropic or Gemini SDK: start with Intent Gateway. It takes a base URL change.
- Your agent runs in a sandbox and acts on files, shells and local sockets that never pass through an API: look at Intent Container.
- You want enforcement inside your own code, per step: use the SDK.