Get started
The recommended way to start is to connect your coding agent to the Tracecat MCP server. Your agent then builds and runs workflows, agents, tables, and cases for you.Getting started
Connect your coding agent to Tracecat Cloud or a self-hosted instance.
Tracecat MCP
View every tool your agent can call in Tracecat.
Deploy Tracecat
Use Tracecat Cloud or self-host Tracecat in your own infrastructure.Tracecat Cloud
Sign up and start building without running any infrastructure.
Docker Compose
Run every Tracecat service on a single host in minutes.
AWS ECS Fargate
Deploy to your own AWS account with Terraform.
Kubernetes
Install the Helm chart on EKS, GKE, or any conformant cluster.
Air-gapped
Run on Kubernetes with no runtime internet access.
Architecture
See how Tracecat services fit together before you deploy.
Explore the platform
Agents
Build custom AI agents that call tools and MCP (stdio, remote) servers.
Workflows
Draft, publish, and run workflow definitions.
Actions
Configure action inputs, success and error paths, and control flow.
Triggers
Start workflows from external systems with webhooks.
Case management
Resolve work items with agents and workflows.
Workspace chat
Chat across your workspace with tools, MCP servers, and preset agents.
Tables
Store structured data and query it from workflows and agents.
Integrations
Configure prebuilt credentials for Tracecat’s built-in integrations.
Custom integrations
Build your own custom actions and sync them into Tracecat via git.
MCP servers
Connect third-party remote and stdio MCP servers for your agents to call.