Grok and Dome
Run Grok 4.7 from xAI, with a fallback, under one name. Grok often starts as one team's request, well before it's anyone's standard. Dome lets research have it without opening it to the company: xAI serves it, OpenRouter backs it up, a rule checks the person behind each call, and a monthly cap bounds the trial.
Dome provides model brokeringand routing for Grok.
xAI first, OpenRouter next
OpenRouter serves Grok as x-ai/grok-4.7. Pool it behind xAI and a failed call moves to it.
Grok for research
Tag Grok connections with a family attribute. A rule refuses them unless the person the agent acts for is in research.
A ceiling on spend
$200 of provider spend a month covers the trial. Calls over the line are refused, whichever provider would have served them.
Grok with failoverin five steps.
xAI, then OpenRouter, in one pool, opened to research and capped at $200.
Connect xAI
The provider id is xai. Dome holds the xAI key, and the research agent calls with its own.
shelldome models add grok-xai \ --provider xai \ --model grok-4.7 \ --api-key "$XAI_API_KEY" \ --attributes '{"family":"grok"}' \ --gateway prod-gatewayConnect OpenRouter
OpenRouter names the model with an x-ai/ prefix.
shelldome models add grok-openrouter \ --provider openrouter \ --model x-ai/grok-4.7 \ --api-key "$OPENROUTER_API_KEY" \ --attributes '{"family":"grok"}' \ --gateway prod-gatewayPool them for failover
xAI is at priority 0 and OpenRouter at 1, and the research agent sends grok-4.7 either way.
shelldome models pool create grok-4.7 \ --failover-max all --gateway prod-gateway dome models pool member add grok-4.7 grok-xai --priority 0dome models pool member add grok-4.7 grok-openrouter --priority 1Point your agent at the Gateway
xAI's API is OpenAI-compatible, so the OpenAI SDK needs only the Gateway URL and a Dome key.
pythonfrom openai import OpenAI client = OpenAI(base_url=f"{GATEWAY_URL}/v1", api_key=DOME_AGENT_KEY)client.chat.completions.create( model="grok-4.7", messages=[{"role": "user", "content": "Summarize this week's chip export news."}],)Cap the spend
One quota on the pool covers xAI and OpenRouter together.
shelldome quotas set --subject pool --pool grok-4.7 \ --unit provider_usd --limit 200 --window monthly
Grokfor the research team.
On an agent with --agent, Grok is refused unless the person behind the call is in research. The agent's other models keep working.
cedar
forbid (principal, action == Dome::Action::"llm:invoke", resource is Dome::LLMModel)when { resource has family && resource.family == "grok" }unless { principal has act_as && principal.act_as.groups.contains("research") };One call,two outcomes.
Switch the caller or the argument and watch the same call decide differently, with every decision landing in audit.
Acting for
agent research-agent · acting as a.park
llm:invoke(model: "grok-4.7")
- Agentresearch-agent is registered and active
- Callera.park verified, groups: research
- RuleGrok is open to the research group
- Quota$74 of $200 this month
DecisionAllowed
Bringing it alltogether.
Connecting Grok to registered agents, tools, and identity in Dome completes a governed agent application.
Acting for
Control point
GatewayEvery call decided and audited
RulesGuardsQuotas
Models
Common questionsabout Grok.
More of whatDome works with.
Provider
Anthropic
The Claude API behind the Model Broker: the key held in Dome, every call authorized, metered and audited.
MCP server
GitHub
The GitHub MCP server behind the Tool Gateway, with rules that decide each call on its owner and repository.
Identity
Microsoft Entra ID
Entra access tokens verified on every agent call, so rules read app roles and audit names the person.
Runtime
LangGraph
LangGraph agents with their model calls on the Model Broker and their MCP tools on the Tool Gateway.
Client
Claude Code
Claude Code on a Dome Gateway with per-developer sign-in, rules on every tool call, and audit by name.
Agent service
TinyFish
TinyFish's web agents behind the Tool Gateway, with rules that decide each run on the site it targets.
Talk with ourFDE team.
Or start free, and register your first agent in minutes. Connect it through a gateway, give it rules, and see every call it makes in one audit trail.
Talk with our FDE team
Our engineers design, build and run governed agents for the part of your business you want to change. You see what they deliver.
AI models for enterprise agents: failover, rules and quotas
Every model your agents call goes through the Model Broker. Pool providers for failover, decide who may call each model, and cap what it costs.