> ## Documentation Index
> Fetch the complete documentation index at: https://browseruse-0aece648-codex-deterministic-rerun-docs.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Hermes Agent

> Give Hermes Agent cloud browser automation with Browser Use.

[Hermes Agent](https://github.com/nousresearch/hermes-agent) is an open-source, self-improving AI agent by Nous Research. It has built-in browser automation tools that work with local Chromium out of the box. Add Browser Use and those tools run on cloud browsers with anti-detect profiles, residential proxies in 195+ countries, and stealth browsing.

Two ways to set it up: configure Browser Use as Hermes's cloud browser backend, or install the Browser Use CLI and let Hermes drive it directly.

## Option 1: Cloud Browser Backend

Hermes has built-in browser tools (`browser_navigate`, `browser_click`, `browser_snapshot`, etc.) that default to local Chromium. Point them at Browser Use cloud browsers instead — no extra dependencies, same Hermes experience.

### Setup

**1. Get your API key**

Sign up at [cloud.browser-use.com](https://cloud.browser-use.com) and copy your API key from [Settings → API Keys](https://cloud.browser-use.com/settings?tab=api-keys\&new=1).

Or let the agent provision one itself — see [Agent Self-Registration](#agent-self-registration) below.

**2. Configure Hermes**

Run the setup wizard:

```bash theme={null}
hermes setup tools
```

Select **Browser Automation**, then **Browser Use**, and paste your API key when prompted.

Or configure manually — add your key to `~/.hermes/.env`:

```bash theme={null}
BROWSER_USE_API_KEY=your_key_here
```

And set the provider in `~/.hermes/config.yaml`:

```yaml theme={null}
browser:
  cloud_provider: browser-use
```

**3. Use it**

Just chat with Hermes — any browsing tasks automatically route through Browser Use cloud browsers:

```
> Find the top trending repositories on GitHub today and summarize them
```

## Option 2: Browser Use CLI

The [Browser Use CLI](https://docs.browser-use.com/open-source/browser-use-cli) is a standalone tool that gives Hermes browser automation through terminal commands. Hermes drives the browser directly via its terminal tool — giving you shared browser sessions across agents, persistent logins and cookies, profile management, and access to Browser Use's full command surface.

### Setup

**1. Install the CLI**

```bash theme={null}
curl -fsSL https://browser-use.com/cli/install.sh | bash
```

**2. Verify the installation**

```bash theme={null}
browser-use doctor
```

**3. Install the skill**

The Browser Use skill teaches Hermes the full CLI command set. Install it from [skills.sh](https://skills.sh/browser-use/browser-use/browser-use):

```bash theme={null}
hermes skills install skills-sh/browser-use/browser-use/browser-use
```

Or ask Hermes directly in chat to install it.

**4. Connect to cloud browsers**

Log in with your API key:

```bash theme={null}
browser-use cloud login <your-api-key>
```

Or let the agent provision one itself — see [Agent Self-Registration](#agent-self-registration) below.

**5. Use it**

Once the skill is loaded, Hermes can drive the browser through CLI commands via its terminal tool:

```
> Use browser-use to open github.com/trending and summarize the top repos
```

For the complete CLI reference and advanced features like cloud browsers, tunnels, sessions, and Python execution, see the [Browser Use CLI docs](https://docs.browser-use.com/open-source/browser-use-cli).

## Agent Self-Registration

Hermes can provision its own Browser Use API key autonomously — no human interaction needed. This works with both options above.

Install the Browser Use CLI:

```bash theme={null}
curl -fsSL https://browser-use.com/cli/install.sh | bash
```

The agent runs three CLI commands using its terminal tool:

**1. Request a challenge**

```bash theme={null}
browser-use cloud signup
```

Returns a `Challenge ID` and `Challenge` — an obfuscated math problem that requires LLM reasoning to solve.

**2. Solve the challenge and verify**

The agent solves the math problem, then verifies with the answer (a string with 2 decimal places, e.g. `"13.60"`):

```bash theme={null}
browser-use cloud signup --verify <challenge-id> <answer>
```

The CLI saves the API key to `~/.browser-use/config.json` automatically.

**3. Copy the key to Hermes config**

For the cloud browser backend (Option 1):

```bash theme={null}
hermes config set BROWSER_USE_API_KEY $(browser-use config get api_key)
```

For CLI mode (Option 2), the key is already saved — no extra step needed.

### Claim the account (optional)

If the human wants to see the account in the dashboard later:

```bash theme={null}
browser-use cloud signup --claim
```

Returns a claim URL valid for 1 hour. The human can follow the link to sign up and claim the agent-created account.
