---
title: "Monitor Usage and Peer Sessions in Autohand Code Docs"
source: https://docs.autohand.ai/guides/usage-and-peers
---

# Monitor usage and peer sessions

Sign in to choose a hosted or local model, use /usage to understand your account and project activity, and use /peers to see the other active sessions in this workspace.

## 1\. Sign in and choose a model

Start Autohand Code in a repository, then use `/login`. It opens the device sign-in flow in your browser. After you return to the terminal, use `/model` whenever you want to inspect or change the active model.

Autohand Code

```bash
autohand

# In the Autohand prompt
/login
/model
```

For a profile that has never chosen a provider, signing in selects hosted **Fantail**. An explicit provider selection is preserved. In `/model`, choose hosted **Fantail** for quick, focused coding work (256K context), hosted **Moa** for deliberate work across more context (1M context), or the Local path on a Mac with Apple Silicon to configure a compatible MLX model.

See [Autohand models](https://docs.autohand.ai/models/) for the complete model guide and [Autohand Console](https://docs.autohand.ai/getting-started/console) for the account view.

## 2\. Read usage in the terminal

`/usage` combines the account limits available to the selected Autohand account with activity from the current Code client. It is the fastest way to check the coding history that matters to the project you are working in.

Autohand prompt

```bash
/usage
/usage daily
/usage weekly
/usage monthly
```

-   `/usage` and `/usage daily` show daily activity for the past 12 months.
-   `/usage weekly` groups activity into the past 52 weeks.
-   `/usage monthly` shows the past 12 months by month.

When you use hosted Autohand, the view includes the plan and currently enforceable allowance or throughput windows, as well as activity totals, peak days, and streaks. Reset times are formatted for the CLI environment. With another provider, the terminal can still show local activity but may not have Autohand account entitlement data.

**Account-wide view:** Open [Console → Usage](https://console.autohand.ai/usage) for the account-level view. Use it alongside `/usage`, rather than expecting a project-local terminal session to be an organization-wide ledger.

## 3\. See concurrent sessions with `/peers`

Use `/peers` in a workspace that may have more than one active Autohand Code session. It is a read-only awareness tool: it does not start agents, stop agents, or send messages.

Autohand prompt

```bash
/peers
```

The result identifies active sessions and, where available, shows their model, provider, mode, current status, context use, start time, activity phase, branch head, and the paths or claims they have reported. In an interactive terminal it opens a scrollable view; in a non-interactive terminal it returns text cards.

**No peers is a useful answer.** If no other session is active, Autohand says so. If peer awareness is unavailable in the current runtime, it reports that instead. Neither response changes your workspace.

## Use the three views together

| Question | Use |
|---|---|
| What is my account allowance and recent account usage? | Console → Usage |
| What has this Code client done over time? | /usage, /usage weekly, or /usage monthly |
| Who else is working in this workspace right now? | /peers |

## Send messages to local peers

With `sessions.communication.enabled`, use the colon composer to select an exact root or worker, send directly, and inspect delivery through `/peers inbox`. Follow the [peer communication guide](https://docs.autohand.ai/guides/peer-communication), the [two-session lab](https://docs.autohand.ai/tutorials/peer-communication-lab), or the [technical reference](https://docs.autohand.ai/guides/peer-communication-protocol).

For shared build capacity, see [resource coordination](https://docs.autohand.ai/guides/peer-resource-coordination). Controller grants are enforced before participating commands start, while existing builds retain ownership until their process trees finish.

## Next steps

-   Read the [slash command reference](https://docs.autohand.ai/working-with-autohand-code/slash-commands) for all interactive commands.
-   Use [Autohand Console](https://docs.autohand.ai/getting-started/console) to manage API keys, Code clients, memories, connectors, and account settings.
-   Create an API key for an application with the [Code Agent SDK tutorial](https://docs.autohand.ai/agent-sdk/api-key-setup).