Frequently asked questions

Pricing is shared on the discovery call, once we understand your workflow and what you want your assistant to own. It's a done-for-you setup, scoped to your business, not a one-size-fits-all plan.

Book a discovery call. We spend 30 minutes mapping what you'd hand off, confirm your tools are connectable, and outline exactly what your assistant will run.

Solo founders and lean teams who need an executive assistant but don't want to hire, onboard, and manage one. If your day is full of triage, scheduling, follow-up, and reporting, this is built for you.

Because a Claw actually does the work instead of just talking about it. A Claw is an AI agent that operates a computer the way a person does: you give it a goal, it breaks that goal into steps, then uses real tools like a web browser, a terminal, or your apps to carry them out. ChatGPT waits for you to prompt it, a VA needs hiring and managing, and Zapier breaks the moment a task needs judgment. Your Claw does all three: it runs on its own, makes judgment calls, and connects your tools, on infrastructure you own, learning your preferences over time.

We believe every person should have their own team of agents, not share one. So rather than a single assistant for the whole company, we recommend each team member runs their own fleet of Claws, tuned to their role, their tools, and their preferences. Everyone gets an assistant that works the way they do, and your team multiplies its output across the board.

A 30-minute discovery call to map what you'll hand off, then most clients are live within days. We handle installation, tool connections, memory, and the rules for what needs your sign-off while you review and approve each piece as it comes online.

Yes. Every engagement includes post-launch support and full handoff documentation. Optional monthly maintenance is available for teams that want ongoing tuning.

Yes. We provide a proper invoice with your company name and details for the engagement.

Most clients have their assistant running in days. We handle installation, tool connections, memory, and approval gates; you review and approve as it comes online.

No. You interact with your assistant the way you'd message a person, via Telegram or Slack. We configure everything during setup.

On infrastructure you own: a secure VPS or your own Mac Mini. There's no shared SaaS backend; your data stays on your server.

Email, calendar, and Slack are the core, plus tools like Stripe, Zoom, and the other services your team already uses. You reach the assistant through Telegram or Slack. If you can connect a tool, your assistant can run it, and we wire it up during setup.

Just the discovery call and access to the tools you want it to run. We configure everything else: installation, connections, memory, and the rules for what needs your sign-off, then hand off documentation when it's live.

Yes. We install on your stack, configure it to your workflow, and hand off complete documentation so your team knows exactly how it works.

AGENTS.md is the prompt + workflow for the agent. TOOLS.md describes the integrations it touches. HEARTBEAT.md is a rolling log it writes to. The three together are a complete operator's manual.

Yes, it's your installation. We hand off full documentation so you (or your team) can adjust workflows, prompts, and automations whenever your business changes.

It runs on your own server and connects to the tools you approve. Your data stays on your infrastructure; nothing routes through us.

No. It's built on an open agent runtime, and your AI API keys go directly from your machine to Anthropic or OpenAI. You own the installation and can move it.

The main “brain” runs on a frontier model, typically something like Claude Opus 4.8 or GPT-5.5, for the hardest reasoning. Behind it, role-specific subagents run on highly capable but more economical models, so you get top-tier judgment where it matters without paying frontier rates for every task.

Yes. You bring your own AI keys, so usage is billed to you at cost with no markup. We recommend the right models for your workload and wire everything up through OpenRouter, which keeps the integration simple and lets you switch providers easily.

It depends on how much work you hand off and which models you choose, but because you pay providers directly through OpenRouter, you only ever pay at cost with no markup. We help you pick models that balance capability and price, and most clients find it's a fraction of what an assistant's salary would be.

Your assistant runs on infrastructure you own, your VPS or Mac Mini, not a shared backend we operate. Conversations, files, and working memory stay on your server, and your AI keys go directly from your machine to the model provider. We don't route your data through us, and memory lives in plain files you can inspect, edit, export, or delete at any time.

It handles routine work on its own, but checks with you before doing anything that matters, like sending an email, spending money, posting something, or making a big call. It only moves once you say yes, and you decide which actions need a thumbs-up first. Give it more freedom or keep it on a tighter leash whenever you want.

Approval gates are checkpoints where the assistant stops and asks for your OK before taking actions that affect customers, money, or external systems. They let you get 80–90% of the time savings from automation while keeping final control over sensitive decisions. Learn more about [approval-gated AI agents](/approval-gated-ai-agents).

Yes. It uses persistent semantic memory, so it learns your preferences and context across every session. With optional maintenance, we also add new capabilities as your needs grow.

It's built on an extensible agent runtime with a skill and runbook system. The handoff docs cover how to add automations yourself, or we can handle it under a maintenance plan.

You keep the assistant. Maintenance covers ongoing support and tuning, not access to the software. Your installation keeps running.

Email support@clawempire.com. Real humans, usually a same-day reply, and we love seeing what people are running.

Reach out and we'll fix it. For the best peace of mind we recommend a maintenance plan, where we proactively monitor your assistant and resolve issues before they reach you, so your agents stay on and working.

Yes. Tell us what you'd like your assistant to handle. We'll scope it and add it, typically under a maintenance plan.

We keep an eye on major providers (Gmail, Slack, Stripe, calendars). Under maintenance, we update connections when their APIs change so your assistant keeps working.

Proactive monitoring so we catch and fix issues before they reach you, updates when providers change their APIs so integrations keep working, ongoing tuning as your workflow evolves, and new automations or capabilities on request. Maintenance is optional. If you stop, you keep your assistant and it keeps running.