Documentation/docs/managed-agents/sdk

#SDK Usage

Use the official Anthropic SDK and point it at the Sandbox0 Managed Agents API endpoint.

SDK field	Sandbox0 value
`baseURL`	`https://agents.sandbox0.ai`
`apiKey`	Sandbox0 API key
`authToken`	Same Sandbox0 API key
Model token	Stored in an LLM vault, not passed to the SDK client

Official reference: Claude Managed Agents quickstart.

Minimal TypeScript Flow#

typescript
import Anthropic from "@anthropic-ai/sdk";

const sandbox0ApiKey = process.env.SANDBOX0_API_KEY;

const client = new Anthropic({
    baseURL: "https://agents.sandbox0.ai",
    apiKey: sandbox0ApiKey,
    authToken: sandbox0ApiKey,
});

const suffix = `${Date.now().toString(36)}-${Math.random().toString(36).slice(2, 8)}`;

const environment = await client.beta.environments.create({
    name: `demo-env-${suffix}`,
    config: {
        type: "cloud",
        networking: { type: "unrestricted" },
        packages: { type: "packages" },
    },
});

const agent = await client.beta.agents.create({
    name: `Demo Agent ${suffix}`,
    model: { id: process.env.MODEL_ID ?? "claude-sonnet-4-20250514" },
    system: "Reply in one short sentence.",
    tools: [{ type: "agent_toolset_20260401" }],
});

const llmVault = await client.beta.vaults.create({
    display_name: `Claude LLM ${suffix}`,
    metadata: {
        "sandbox0.managed_agents.role": "llm",
        "sandbox0.managed_agents.engine": "claude",
        "sandbox0.managed_agents.llm_base_url": process.env.MODEL_BASE_URL ?? "https://api.anthropic.com",
    },
});

await client.beta.vaults.credentials.create(llmVault.id, {
    display_name: "Anthropic-compatible API key",
    auth: {
        type: "static_bearer",
        token: process.env.MODEL_API_KEY!,
    } as any,
});

const session = await client.beta.sessions.create({
    agent: agent.id,
    environment_id: environment.id,
    title: `Sandbox0 demo ${suffix}`,
    vault_ids: [llmVault.id],
});

await client.beta.sessions.events.send(session.id, {
    events: [{
        type: "user.message",
        content: [{ type: "text", text: "Say hello from Sandbox0 managed agents." }],
    }],
});

for await (const event of client.beta.sessions.events.list(session.id, { limit: 100 })) {
    if (event.type === "agent.message") {
        console.log(event.content?.map((block) => block.text ?? "").join(""));
    }
}

Required Variables#

Variable	Purpose
`SANDBOX0_API_KEY`	Authenticates the SDK client to Sandbox0
`MODEL_API_KEY`	Model provider token stored in the LLM vault
`MODEL_BASE_URL`	Optional Anthropic-compatible model endpoint
`MODEL_ID`	Optional model id for the agent

ℹ

Do not set the SDK client's apiKey to the model provider token. That key authenticates to Sandbox0. The model token belongs in the LLM vault.

ℹ

Set both apiKey and authToken to the Sandbox0 API key. The Anthropic SDK can also read ANTHROPIC_AUTH_TOKEN from the process environment; explicitly setting authToken prevents that model-provider token from being sent as the Sandbox0 client bearer token.

ℹ

The as any cast on unbound static_bearer credentials is intentional for current Anthropic TypeScript SDK types. Sandbox0 LLM credentials must omit mcp_server_url.

Direct HTTP Calls#

Prefer the SDK for application code. Direct HTTP calls are useful for debugging.

bash
curl -fsS "https://agents.sandbox0.ai/v1/sessions" \
    -H "x-api-key: $SANDBOX0_API_KEY" \
    -H "anthropic-version: 2023-06-01" \
    -H "anthropic-beta: managed-agents-2026-04-01" \
    -H "content-type: application/json" \
    -d '{
        "agent": "agent_...",
        "environment_id": "env_...",
        "vault_ids": ["vlt_..."]
    }'

#SDK Usage

Minimal TypeScript Flow#

Required Variables#

Direct HTTP Calls#

Next Steps#

Agents

Environments