What is CRHQ?

CRHQ (Control Room HQ) is an autonomous AI agent platform. Each CRHQ instance runs on a dedicated server, giving your AI agents their own workspace — with file system access, browser automation, persistent memory, and the ability to connect to your tools and services.

Think of it as a workspace where AI agents work alongside you. You give them instructions, skills, and access — they research, build, create, and strategize.

Core Concepts

Before diving into the platform, here are the key building blocks:

Sessions

A session is a conversation between you and an agent. You can create multiple sessions with different agents, organize them into projects, and come back to them at any time. Sessions support text, voice input, file attachments, and artifacts.

Agents

Agents are AI workers with specific roles and capabilities. Each agent has a name, instructions that shape its behavior, a default model (Haiku, Sonnet, or Opus), and assigned skills. You can use the pre-built system agents or create your own.

Skills

Skills are capabilities you give to agents. A skill contains instructions, API references, and scripts that tell an agent how to do something — like manage a database, use a browser, connect to Slack, or generate reports. Skills are modular: assign only what each agent needs.

Recipes

Recipes are procedural, multi-step workflows. While skills give agents capabilities, recipes give them procedures — step-by-step instructions for executing economically valuable work. For example: "Use these 7 skills to audit a website, generate a report, and send it via email."

Projects

Projects organize your work. Group sessions, assign project documents, and track status. Each project can have its own knowledge base that agents automatically reference when working in that context.

Memory

Memory is persistent knowledge that agents retain across sessions. Facts, preferences, decisions, and context — once stored, agents can recall and apply this information in future conversations automatically.

How Agents Work

Under the hood, each agent runs as an isolated process on your dedicated server. When you send a message:

1. Your message is queued and delivered to the agent
2. The agent reads its instructions, skills, and relevant context
3. It processes your request — potentially using tools, browsing the web, reading files, or calling APIs
4. Results are streamed back to you in real-time
5. The conversation is saved and summarized automatically

Agents have access to:

• File system — Read, write, and manage files on the server
• Web browser — Autonomously navigate websites, fill forms, take screenshots
• External services — Connect to APIs via credentials and connectors
• Other agents — Delegate tasks to specialized agents
• Scheduled execution — Run jobs on a schedule without human intervention

Your Dedicated Server

Every CRHQ instance runs on its own VPS (Virtual Private Server). This means:

• Privacy — Your data stays on your server
• Performance — No shared resources or noisy neighbors
• Full access — Agents have a real file system, not a sandbox
• Always on — Agents can work 24/7 via background jobs