User Guide on Runabot Documentation/docs/en/user/Recent content in User Guide on Runabot DocumentationHugoenGetting Started/docs/en/user/getting-started/Mon, 01 Jan 0001 00:00:00 +0000/docs/en/user/getting-started/<h1 id="getting-started-with-runabot">Getting Started with Runabot<a class="anchor" href="#getting-started-with-runabot">#</a></h1> <p>Runabot allows you to deploy and manage AI assistants (bots) in an isolated environment.</p> <h2 id="1-login">1. Login<a class="anchor" href="#1-login">#</a></h2> <p>Navigate to the Runabot UI (e.g., <code>https://runabot.de/ui/</code>). If not logged in, you will be redirected to the login page. Use one of the sign-in methods offered by your Runabot environment.</p> <h2 id="2-deploy-a-bot">2. Deploy a Bot<a class="anchor" href="#2-deploy-a-bot">#</a></h2> <ol> <li>On the Dashboard, click <strong>Deploy New Bot</strong>.</li> <li>Enter a name (e.g., <code>my-assistant</code>).</li> <li>Select <strong>Bot Type (Chart)</strong>, for example <code>openclaw</code> or <code>vscode</code>.</li> <li>Click <strong>Deploy</strong>.</li> </ol> <p>The bot will appear in your list. Provisioning can take several minutes. Check its runtime status and deployment status before opening the terminal. If provisioning fails, inspect the bot&rsquo;s events with the <a href="/docs/en/user/cli/">CLI</a>.</p>UI Reference/docs/en/user/ui-reference/Mon, 01 Jan 0001 00:00:00 +0000/docs/en/user/ui-reference/<h1 id="ui-reference">UI Reference<a class="anchor" href="#ui-reference">#</a></h1> <h2 id="dashboard">Dashboard<a class="anchor" href="#dashboard">#</a></h2> <p>The main dashboard provides an overview of your resources.</p> <ul> <li><strong>My Bots</strong>: Grid view of your running bot instances. <ul> <li><strong>Status Indicator</strong>: Green (Running), Orange (Pending), Red (Failed).</li> <li><strong>Delete</strong>: Starts bot deletion. Backup and cleanup progress may delay completion.</li> <li><strong>Terminal</strong>: Open the web terminal.</li> </ul> </li> <li><strong>Deploy New Bot</strong>: Opens the deployment wizard.</li> </ul> <p>Each bot card shows compact uploaded (↑) and downloaded (↓) traffic below its last backup. The <strong>My Addons</strong> overview shows the same figures for each addon. <code>N/A</code> means no measurement is available; it does not mean zero traffic.</p>Budgets, traffic credits, and payment estimates/docs/en/user/billing/Mon, 01 Jan 0001 00:00:00 +0000/docs/en/user/billing/<h1 id="budgets-traffic-credits-and-payment-estimates">Budgets, traffic credits, and payment estimates<a class="anchor" href="#budgets-traffic-credits-and-payment-estimates">#</a></h1> <p>Runabot aims to keep your costs predictable through spending limits and clear payment estimates. Compute and storage are billed differently from network traffic, as explained below.</p> <h2 id="compute-and-storage-limit">Compute and storage limit<a class="anchor" href="#compute-and-storage-limit">#</a></h2> <p>The CPU, RAM, and disk amount on the Usage page is a maximum weekly spending limit. Raising it does not purchase anything by itself. Charges are based on the CPU and RAM allocated to running bots and addons, plus retained storage. Unused headroom is not charged.</p>MCP Proxy Addon/docs/en/user/mcpproxy/Mon, 01 Jan 0001 00:00:00 +0000/docs/en/user/mcpproxy/<h1 id="mcp-proxy-addon">MCP Proxy Addon<a class="anchor" href="#mcp-proxy-addon">#</a></h1> <p>The <strong>MCP Proxy</strong> addon is a secure gateway and mediation layer for Model Context Protocol (MCP) servers in Runabot. It allows AI agents, developer workspaces, and autonomous bots (such as OpenClaw, Codex, Claude Code, or VS Code) to interact with external tools and APIs without holding sensitive third-party credentials directly in the workload environment.</p> <h2 id="why-mcp-proxy">Why MCP Proxy?<a class="anchor" href="#why-mcp-proxy">#</a></h2> <p>In an AI agent environment, granting agents direct access to raw API keys or unrestricted network endpoints introduces serious security risks. MCP Proxy solves this by enforcing a zero-trust model:</p>CLI Reference/docs/en/user/cli/Mon, 01 Jan 0001 00:00:00 +0000/docs/en/user/cli/<h1 id="runabot-cli">Runabot CLI<a class="anchor" href="#runabot-cli">#</a></h1> <p>The <code>runabot</code> command-line tool lets you manage workloads, addons, and account usage from a terminal or script. Use it interactively, automate tasks in CI/CD pipelines, or let your AI agent manage Runabot resources on your behalf.</p> <hr> <h2 id="1-download--installation">1. Download &amp; Installation<a class="anchor" href="#1-download--installation">#</a></h2> <p>You can download pre-built binaries for <strong>Linux</strong>, <strong>macOS</strong>, and <strong>Windows</strong> (supporting both <code>amd64</code> and <code>arm64</code>) directly from the <strong><a href="/docs/downloads/">Downloads Page</a></strong>.</p> <h3 id="quick-install">Quick Install<a class="anchor" href="#quick-install">#</a></h3> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl"><span class="c1"># User-level installation (installs to ~/.local/bin/):</span> </span></span><span class="line"><span class="cl">curl -fsSL __BASE_URL__/downloads/install.sh <span class="p">|</span> sh </span></span><span class="line"><span class="cl"> </span></span><span class="line"><span class="cl"><span class="c1"># Global system installation (installs to /usr/local/bin/):</span> </span></span><span class="line"><span class="cl">curl -fsSL __BASE_URL__/downloads/install.sh <span class="p">|</span> sudo sh</span></span></code></pre></div><p>Or download the binary manually from the <a href="/docs/downloads/">Downloads Page</a>, make it executable, and move it into your <code>PATH</code>:</p><link>/docs/en/user/openclaw/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>/docs/en/user/openclaw/</guid><description><h1 id="openclaw-bot-guide">OpenClaw Bot Guide<a class="anchor" href="#openclaw-bot-guide">#</a></h1> <p>OpenClaw is a powerful personal assistant that you can run on Runabot. This guide covers common administrative tasks for your OpenClaw instance.</p> <h2 id="initial-setup-onboarding">Initial Setup (Onboarding)<a class="anchor" href="#initial-setup-onboarding">#</a></h2> <p>When you first deploy an OpenClaw bot, you need to perform an initial onboarding to set up your administrator account and basic settings.</p> <ol> <li> <p>Open the <strong>Web Terminal</strong> for your bot.</p> </li> <li> <p>Run the onboarding command:</p> <div class="highlight"><pre tabindex="0" class="chroma"><code class="language-bash" data-lang="bash"><span class="line"><span class="cl">openclaw onboard</span></span></code></pre></div></li> <li> <p>Follow the interactive prompts to:</p> <ul> <li>Set up your administrator username and password.</li> <li>Configure basic system settings.</li> <li>Initialize the local database.</li> </ul> </li> </ol> <p>Once onboarding is complete, you can access the OpenClaw Web UI using the credentials you just created.</p></description></item><item><title/><link>/docs/en/user/security/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>/docs/en/user/security/</guid><description><h1 id="security-features">Security Features<a class="anchor" href="#security-features">#</a></h1> <p>Runabot&rsquo;s security model is based on zero trust for AI agents and development workloads: no real credentials in the runtime, no default external network access, and every important tool boundary mediated by a proxy.</p> <p>This page describes the user-facing security controls. For a broader product overview, see <a href="/docs/en/user/zero-trust-runtime/">Zero-Trust Agent Runtime</a>.</p> <h2 id="runtime-boundary-overview">Runtime Boundary Overview<a class="anchor" href="#runtime-boundary-overview">#</a></h2> <p>The workload can be a coding agent, a VS Code server, a web terminal, or another developer tool running inside the Runabot-managed container. It does not connect directly to external systems. Each channel is routed through a Runabot-managed proxy, addon, or policy boundary first, where policy and credentials are enforced before traffic can leave the environment.</p></description></item><item><title/><link>/docs/en/user/workloads/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>/docs/en/user/workloads/</guid><description><h1 id="workload-types">Workload Types<a class="anchor" href="#workload-types">#</a></h1> <p>Runabot is not limited to one bot implementation. It provides confined runtimes for agents, developer workspaces, and automation jobs that need useful access without direct exposure to real credentials.</p> <h2 id="openclaw-ironclaw-hermes">OpenClaw, IronClaw, Hermes<a class="anchor" href="#openclaw-ironclaw-hermes">#</a></h2> <p>OpenClaw, IronClaw and Hermes are the ready-to-run autonomous assistants for users who want a complete agent experience out of the box. It is a good fit when the workload is primarily an AI assistant with its own UI and long-running state.</p></description></item><item><title/><link>/docs/en/user/zero-trust-runtime/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>/docs/en/user/zero-trust-runtime/</guid><description><h1 id="zero-trust-agent-runtime">Zero-Trust Agent Runtime<a class="anchor" href="#zero-trust-agent-runtime">#</a></h1> <p>Runabot is a zero-trust runtime for AI agents, developer workspaces, CI/CD jobs, and package build environments. It deploys and manages useful workloads while keeping real credentials and broad network access out of the runtime.</p> <p>The runtime is meant for AI agents, developer workspaces, CI/CD jobs, and package build environments that need external dependencies without exposing real tokens.</p> <h2 id="design-goal">Design Goal<a class="anchor" href="#design-goal">#</a></h2> <p>Runabot assumes that agents will read hostile content, run imperfect tools, and sometimes be manipulated by prompt injection or compromised dependencies.</p></description></item></channel></rss>