Model proxy
Also written LLM gateway
A service between an application or agent and its model provider's API: the client points its base URL at the proxy, which forwards each request and can route, rate-limit, log, redact or refuse it. It sees prompts, tool definitions and the tool calls the model returns, but not what then happens on the machine, so it can refuse a tool call in a response yet cannot observe the local command that call becomes.
How MoorAI handles it: moorai-model-proxy sits between an agent's model SDK and Anthropic, OpenAI, Gemini or a local Ollama, scans what the agent sends and decides each tool call the model returns.
See also MCP gateway · Inference hooks · action layer
Related terms
-
MCP gateway
Agents & architecture
A proxy between agents and MCP servers that tool listings and tool calls pass through, so one place can authenticate, allow or refuse servers and tools…
-
Inference hooks
Agents & architecture
A Claude Enterprise feature, in beta, that sends each governed request to the organization's own AI security server before inference runs and waits for an…
-
Action layer
Agents & architecture
The enforcement surface at the tool call, as opposed to the prompt layer, which tries to classify text before the model sees it.
Agents & architecture
Where the attacks above actually land. The vocabulary here is worth precision, because the difference between "the model decided" and "the tool ran" is the difference between a risk and an incident.