FAQ

Frequently asked questions about Agentish.

What is Agentish?

Agentish is a visual, node-based framework for building autonomous AI agent pipelines. You design workflows by connecting LLM calls, tools, and decision logic in a graph editor, then run them against real challenges like CTF competitions.

How does Agentish relate to iCTF?

Agentish powers the International Capture The Flag (iCTF) competition. It provides the environment for participants to build and deploy autonomous agents that compete in security challenges.

What is MCP and why does it matter?

MCP (Model Context Protocol) is a standard for connecting AI models to external tools and data sources. Agentish has first-class MCP integration, so you can easily wire agents to code execution, file systems, APIs, and other capabilities.

Where can I learn more?

Start with the Documentation hub. The Agentic Workflow guide teaches concepts from first principles; the Agentish Framework guide covers the practical details of building workflows in the editor.

Can I run an agentish agent in a sandbox locally?

No, at the moment Agentish is only meant to be used in the iCTF sandbox. However, after the competition we will make the agentish-sandbox environment available to you so you can run your agent locally by exporting the bundle and running it in the agentish sandbox.