Can You Just

Automation / AI team members

An AI worker still has a queue.

Can You Just can represent an externally operated AI worker as a schedulable agent resource. Its assigned work uses capacity, inherits project priority and respects prerequisites in the same planning system as human work.

Understand the model ↓See the worked example →

Continuous availability is not infinite capacity.

Designated agent resources have continuous 24-hour availability, including weekends and organisation closures. A human reviewer keeps their own working calendar. Ordinary full-day resources use their configured days and exceptions unless explicitly designated as agents.

A longer estimate still consumes more time. Competing assignments still need an order. An agent finishing overnight does not make a human reviewer available overnight.

Human dependencies remain real.

In the fictional example, Atlas prepares work overnight and Alice reviews it during her next working day. If Atlas’s estimate grows far enough, Alice’s review moves too.

Connected agents can request linked human assistance. A blocking request makes the originating task wait; completing all its blockers allows it to become eligible again, subject to the rest of the plan.

CYJ plans the work. Your worker executes it.

The existing agent API supports connected systems, but Can You Just does not supply, wake or run an AI coding agent. Resource planning and external execution are separate responsibilities.

Follow how agent capacity, priorities and human prerequisites interact in the plan. The public API and integration overviews explain the supported connections; API v1 developer documentation explains reliable worker workflows.

A FICTIONAL SCHEDULING STORY

An overnight worker. A daytime reviewer.

Calculated with the CYJ scheduler

Before

Atlas, an externally operated AI resource, has eight hours of preparation from Monday 17:00 to Tuesday 01:00. Alice reviews at 09:00–10:00.

The change

The preparation estimate grows from eight to eighteen working hours. Availability stays the same.

Before

Project completion: Tue 10:00

Alice
  • Human reviewTue 09:00–Tue 10:00
Atlas (AI)
  • Prepare candidateMon 17:00–Tue 01:00

After

Project completion: Tue 12:00

Alice
  • Human reviewTue 11:00–Tue 12:00
Atlas (AI)
  • Prepare candidateMon 17:00–Tue 11:00

Example assumptions. 19–20 October 2026. Atlas is explicitly designated as an agent with continuous capacity. Alice works weekdays 09:00–17:00. CYJ calculates a plan; it does not execute the AI work.

THE CONSEQUENCE

Atlas’s preparation now finishes Tuesday at 11:00. Alice is available from 09:00, but the dependency keeps her review waiting.

THE REVISED PLAN

Alice reviews from 11:00 to noon. Continuous AI availability and the human calendar remain distinct; more effort still has a consequence.

Explore the model:AI team members →Team scheduling →Dependencies →

Connect the worker to its work.

The assigned-worker API lets an external worker report progress and request human help. Integrations connect source work to the same planning model.

CONNECTED IDEAS

The rest of the story.

Team scheduling Team scheduling determines when assigned resources have productive time. Working patterns, breaks and recorded non-working periods shape the windows that work can use.Priorities Priority controls how eligible planned work competes for limited resource time. Can You Just uses four levels: Critical, High, Normal and Low. Stages inherit their project’s priority.Dependencies A dependency says that work must wait for other work to finish. Can You Just combines those prerequisites with calendars, capacity and priorities when calculating the forecast.

Work, time and consequences.

Explore the connected product ↗
Register your interestComing in a later release · registration is not open yet