3 posts tagged with "agent"

The Toolpack SDK workflow engine is powerful out of the box, but its real strength emerges when you customize it. Instead of relying solely on the built-in AGENT_WORKFLOW, CODING_WORKFLOW, or CHAT_WORKFLOW presets, you can craft workflows that match your exact requirements—whether that's a research agent, a code reviewer, or a deployment orchestrator.

This guide shows you how to build custom workflows from scratch, tune planning behavior, control step execution, and design workflows that fit your specific use cases.

An AI agent that can read files, run shell commands, and hit external APIs is useful. An AI agent that does all of that when you only asked it to answer a question is a problem.

Modes are how Toolpack SDK draws that line. A mode defines three things: which tools the AI can access, what system prompt shapes its behavior, and whether the workflow engine (planning, step execution, retries) is active. You set the mode once at init or switch it at runtime, and the SDK enforces the constraints on every request.

Toolpack SDK ships with a workflow engine that turns a basic AI chat into a goal-driven agent. Instead of responding once and waiting, it plans a series of steps, executes them one by one, handles failures, and reports progress the whole time.

This guide walks you through both workflow types, how to configure them, how to listen to events, and patterns you can use in real applications.