> ## Documentation Index
> Fetch the complete documentation index at: https://docs.compose.market/llms.txt
> Use this file to discover all available pages before exploring further.

# Mode, Sandbox, and Proof

> Control execution mode, isolation, and proof-of-execution with /mode, /sandbox, and /proof.

Three commands control *how* the agent executes, not *what* it does. Use them to switch between solo and swarm, force isolated runs, and require auditable proof.

## /mode — solo vs swarm

By default an agent runs in **solo** mode — it handles everything itself. **Swarm** mode lets the agent delegate work to other deployed agents, enabling multi-agent collaboration on complex tasks.

```typescript theme={null}
// Single agent, no delegation
sdk.agent.stream({ agentWallet, message: "/mode solo", threadId: "t1" });

// Allow delegation to sub-agents
sdk.agent.stream({ agentWallet, message: "/mode swarm", threadId: "t1" });
```

In swarm mode the stream includes child-agent events (`swarm_child_start`, `swarm_child_delta`, `swarm_child_done`) for each delegated task. See [Stream Protocol](/sdk/streaming/protocol) for details.

## /sandbox — isolated execution

The sandbox command forces the agent to run tasks in an isolated environment. This is useful when the agent processes untrusted input or runs code that should not touch the host.

```typescript theme={null}
// Force isolated execution
sdk.agent.stream({ agentWallet, message: "/sandbox on", threadId: "t1" });

// Allow in-process execution
sdk.agent.stream({ agentWallet, message: "/sandbox off", threadId: "t1" });

// Check the current setting
sdk.agent.stream({ agentWallet, message: "/sandbox status", threadId: "t1" });
```

When sandbox is on, every task runs in isolation. If a sandbox is unavailable, the agent reports an error rather than silently falling back to in-process execution.

## /proof — auditable execution

Proof mode makes the agent collect a verifiable bundle for each run — input and output hashes per step, inference run IDs, and optional sandbox metadata. Turn it on when you need an audit trail.

```typescript theme={null}
// Require proof-of-execution
sdk.agent.stream({ agentWallet, message: "/proof on", threadId: "t1" });

// Disable proof collection
sdk.agent.stream({ agentWallet, message: "/proof off", threadId: "t1" });

// Check the current setting
sdk.agent.stream({ agentWallet, message: "/proof status", threadId: "t1" });
```

Proof bundles are hash artifacts. They let you verify that step inputs and outputs match what the agent claims, but they are not trusted-execution-environment attestations.

## History commands

Three commands list past activity for the current thread. Each returns a plain text response in the stream — no action events.

### /thread — past conversations

```typescript theme={null}
sdk.agent.stream({ agentWallet, message: "/thread", threadId: "t1" });
```

Lists previous conversation threads associated with this agent.

### /artifacts — generated media

```typescript theme={null}
sdk.agent.stream({ agentWallet, message: "/artifacts", threadId: "t1" });
```

Lists media the agent has generated — images, audio, video — along with references to retrieve them.

### /receipt — payment receipts

```typescript theme={null}
sdk.agent.stream({ agentWallet, message: "/receipt", threadId: "t1" });
```

Lists x402 payment receipts for the thread, including amounts and settlement details. See [Payments](/sdk/payments/overview) for the receipt data model.

## Combining settings

You can stack these commands. A common secure configuration looks like:

```typescript theme={null}
sdk.agent.stream({ agentWallet, message: "/mode swarm", threadId: "t1" });
sdk.agent.stream({ agentWallet, message: "/sandbox on", threadId: "t1" });
sdk.agent.stream({ agentWallet, message: "/proof on", threadId: "t1" });

// Now send the actual task
sdk.agent.stream({
  agentWallet,
  message: "Audit this codebase and report security issues",
  threadId: "t1",
});
```

## Related

* [Command Reference](/sdk/commands/reference) — every command at a glance
* [Plan](/sdk/commands/plan) — run multi-step plans with tracked tasks
* [Stream Protocol](/sdk/streaming/protocol) — swarm child and action event shapes