Core Concepts

This section explains the fundamental concepts that power FixPanic. Understanding these concepts will help you use the platform more effectively.

Platform Overview

FixPanic is built around three core concepts:

How FixPanic Works

1. Create a Cluster - Clusters act as supervisor agents that manage groups of agents. You can communicate with a "cluster" agent to delegate complex tasks and investigations to its subagents.
2. Deploy Agents - Use the FixPanic CLI tool to easily connect to your servers via SSH and deploy agents you've designed on the dashboard. Important: The system enforces a one-agent-per-server policy; you cannot deploy the same agent to multiple servers. Once deployed, agents are ready to communicate via the chat interface.
3. Interact via Chat - The Chat interface allows you to communicate with both direct agents (agents inside a cluster) and the cluster supervisor (the manager of the agents).
4. Agents Execute Commands - Agents use the terminal to execute commands. FixPanic prevents unsafe or insecure commands and implements "human-in-the-loop" logic to ask for approval when sensitive commands are requested.

Data Flow

1. You send a message through the Chat interface
2. FixPanic routes the message to the appropriate agent
3. The agent executes commands on your server
4. Results stream back through the platform to your chat

Key Principles

AI-Powered Intelligence

Agents understand natural language requests:

• "What's causing high CPU usage?"
• "Show me recent error logs"
• "Is the database connection working?"

They translate these into appropriate system commands.

Human-in-the-Loop Control

You maintain control through:

• Approvals - Review and approve sensitive commands
• Whitelisting - Define what commands can auto-execute
• Audit Logs - Track everything that happens

Security by Design

Security is built into every layer:

• Command Whitelisting - Only safe commands allowed
• Path Restrictions - Sensitive directories blocked
• Resource Limits - CPU, memory, timeout constraints
• Audit Trail - Complete execution history

Core Concept Pages

Platform Components

CLI Tool

The command-line interface for:

• Installing and managing agents
• Deploying to servers
• Local agent management

Learn more: CLI Reference

Dashboard

The web interface for:

• Creating and managing clusters
• Monitoring agent status
• Billing and team management

Learn more: Dashboard Guide

Chat Interface

The AI command center for:

• Natural language interaction
• Real-time command execution
• Approval workflows

Learn more: Chat Interface

Architecture Deep Dive

Agent Architecture

Each agent runs as a lightweight service:

Platform Architecture

The FixPanic cloud platform:

Next Steps

Dive deeper into each concept:

• Agents - Understand agent capabilities and lifecycle.
• Clusters - Learn about infrastructure organization.
• Security - Explore the security model in depth.