Announcing Dome Managed Agents.

Try it now

Moonshot AI and Dome

Trial Kimi with one team, with every call governed. Kimi K3 is the default when you pick Moonshot AI in Dome's model picker. Connect it, tag the connection stage: trial, and only the team running the evaluation gets through. Everyone else's call is refused and audited.

Dome provides model brokeringand routing for Moonshot AI.

Trial access, not key access

The team trying Kimi uses Dome keys. Moonshot's key stays in Dome, so ending the trial never means rotating it.

Kimi models in the picker

Kimi K3, Kimi K2.7 Code and Kimi K2.6 are listed in Dome's model picker. Any other model id Moonshot accepts works too.

A trial for one team

Tag the connection stage: trial. A rule lets only the team running the evaluation use it.

Connect Moonshot AIin three steps.

Add Kimi K3 from Moonshot, pool it with Fireworks AI's copy, and move agents onto the Gateway.
  1. Add the connections

    The provider id is moonshot. Dome uses https://api.moonshot.ai/v1 and Moonshot's own model ids; Fireworks uses its own.

    shell
    dome models add kimi-moonshot \  --provider moonshot \  --model kimi-k3 \  --api-key "$MOONSHOT_API_KEY" \  --attributes '{"stage":"trial"}' \  --gateway prod-gateway
    dome models add kimi-fireworks \  --provider fireworks \  --model accounts/fireworks/models/kimi-k3 \  --api-key "$FIREWORKS_API_KEY" \  --attributes '{"stage":"trial"}' \  --gateway prod-gateway
  2. Pool them

    Moonshot answers first and Fireworks second, and the evaluation agents call the pool, kimi-k3.

    shell
    dome models pool create kimi-k3 \  --failover-max all --gateway prod-gateway
    dome models pool member add kimi-k3 kimi-moonshot --priority 0dome models pool member add kimi-k3 kimi-fireworks --priority 1
  3. Point your agent at the Gateway

    Code written for Moonshot's API runs against the Gateway once it carries the agent's Dome key.

    python
    from openai import OpenAI
    client = OpenAI(base_url=f"{GATEWAY_URL}/v1", api_key=DOME_AGENT_KEY)client.chat.completions.create(    model="kimi-k3",    messages=[{"role": "user", "content": "Write tests for this module."}],)

Trial models for theevaluation team only.

Connections tagged stage: trial refuse every call unless the person the agent acts for is in model-evals. Drop the tag when the trial ends.
cedar
forbid (principal, action == Dome::Action::"llm:invoke", resource is Dome::LLMModel)when { resource has stage && resource.stage == "trial" }unless {  principal has act_as &&  principal.act_as.groups.contains("model-evals")};

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 eval-runner · acting as a.novak
llm:invoke(model: "kimi-k3")
  1. Agenteval-runner is registered and active
  2. Callera.novak verified, groups: model-evals
  3. RuleTrial connections are open to model-evals
DecisionAllowed

Bringing it alltogether.

Connecting Moonshot AI to registered agents, tools, and identity in Dome completes a governed agent application.
Dome

Control point

GatewayEvery call decided and audited
RulesGuardsQuotas

Common questionsabout Moonshot AI.

How does Dome connect to Moonshot AI?
Through Moonshot's OpenAI-compatible API at https://api.moonshot.ai/v1. Dome attaches the Moonshot key to each request as a bearer header.
Which Kimi models can I use?
Any model id Moonshot accepts. Dome's model picker lists Kimi K3, Kimi K2.7 Code and Kimi K2.6.
Can I use Kimi K3 from another provider?
Yes. Fireworks serves it as accounts/fireworks/models/kimi-k3 and Together as moonshotai/Kimi-K3, and a pool can put either behind Moonshot.

More of whatDome works with.

Model

Claude Fable

Fable 5.1 from Anthropic and Amazon Bedrock in one failover pool, open to one group and capped by quota.
MCP server

GitHub

The GitHub MCP server behind the Tool Gateway, with rules that decide each call on its owner and repository.
Identity

Okta

Okta tokens verified on every agent call, so rules and audit name the person each agent acted for.
Runtime

LangGraph

LangGraph agents with their model calls on the Model Broker and their MCP tools on the Tool Gateway.
Client

Codex

Codex 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.

LLM providers for AI agents: one governed path to every model

Connect a provider once. Its key stays in Dome, and every agent call to it is authorized, metered and audited.

Every agent you run, under one system of control.