Multi-step tasks

We're releasing the first major addition to the Harbor task format: multi-step tasks.

A multi-step task runs an agent through a sequence of ordered steps against a single, shared environment. Each step has its own instruction and verifier.

Multi-step tasks are helpful when implementing long-horizon tasks with early stopping conditions, testing continual learning methods like memory, and observing an agent's ability to build on its prior work.

Install Harbor 0.5.0 or newer to use multi-step tasks:

uv tool install "harbor>=0.5.0"

Learn more in the multi-step tasks documentation, or have your coding agent guide you through the process of creating a multi-step task by downloading our create-task skill:

npx skills add harbor-framework/harbor --skill create-task

We're excited to see the multi-step tasks Harbor users build!