← All terms

Task Workflow

An ordered pipeline that chains multiple agent tasks together, possibly across different agents.

A task workflow is an ordered pipeline of agent tasks in Agenhood, where the output of one task becomes the input to the next, and the tasks in the chain can belong to different agents.

How it works in Agenhood

A workflow is built as a sequence of steps, each step being an agent task assigned to a particular agent. When a step completes, its output is passed as input to the next step, so a workflow can, for example, have a research agent gather information and then hand that output to a writing agent that produces a document from it. Agenhood shows workflows as a visual pipeline in the console, and each run of a workflow is tracked with per-run success metrics, so a user can see how often a given pipeline completes successfully across its steps.

Why it matters

Many real tasks are naturally broken into stages that call for different skills, tools, or context, and a single agent task is not always the right unit of work for that. Chaining tasks into a pipeline lets each stage stay focused and independently inspectable, while still composing into a larger result. Because each step is a normal agent task, it can use a schema validated output contract so the next step receives data in a predictable shape instead of having to parse free text.

Cron Schedule vs Task Workflow

A task workflow defines what runs and in what order: a sequence of tasks, potentially across multiple agents, where outputs feed forward. A cron schedule defines when something runs: a cadence, whether one-time or recurring, on which a prompt or an existing workflow is triggered. The two are complementary rather than overlapping; a cron schedule can be the trigger for a task workflow, but a workflow itself has no built-in concept of timing.

Related concepts

Chaining discrete units of work into a directed pipeline, where each stage's output feeds the next stage's input, is a long-standing pattern in general workflow and data pipeline infrastructure; Agenhood applies the same idea to agent tasks rather than build steps or data transformations, and a workflow can itself be the thing a cron schedule triggers, combining ordering and timing without merging the two concepts.

Get started

Deploy your fleet.

Put a fleet of sandboxed agents to work on your own infrastructure, provisioned in seconds and watched live from one console.

Get started

Admin-provisioned · Self-host in one command · Your data never leaves your VM