Task Enrichment Agent (TEA)
Background AI enrichment that helps fill task context and schedule work automatically.
Overview
TEA is Fokus's background task enrichment system. Unlike ODA, you do not open TEA from a dedicated workspace and review a batch of proposals first. Instead, TEA runs after task creation and helps Fokus infer useful task details behind the scenes.
In the current interface, the clearest place to notice TEA is inside the Planner views. Tasks can appear with the AI label and may gain extra context such as priority, energy level, bucket assignment, dates, estimated time, or automatic placement on the timeline when Fokus has enough information to make a confident decision.
TEA works in the background
TEA is currently surfaced as behavior rather than as a dedicated screen. You usually notice its results on the task itself, not by opening a separate TEA panel.
Where You Notice TEA
- Create a new task from a planner view such as
Kanban. - Leave some task fields empty and save the task.
- Watch for AI-applied results such as automatic scheduling or metadata badges on the task card.
- Open existing AI-tagged tasks to review and refine anything Fokus filled in for you.
In the reviewed app state, a title-only test task was immediately auto-scheduled after creation, while other AI-tagged tasks in the planner already displayed combinations of priority, energy level, bucket, date, and duration badges.
What TEA Can Help With
- Add structure to tasks that were captured quickly with only a title.
- Infer missing priority and energy signals when similar work already exists.
- Suggest an estimated duration for scheduling.
- Assign a bucket when Fokus can recognize a likely area of work.
- Hand the enriched task off to scheduling so it can be placed into the timeline.
Common Workflow
Capture a task quickly
Add a task without filling every detail manually. This is the situation where TEA is most useful.
Let Fokus process the task in the background
After the task is created, TEA can analyze similar work, bucket patterns, objective context, and other signals to infer missing properties.
Review the AI-applied result in Planner
Look for the AI marker and any added metadata, such as a bucket, effort signal, or scheduled time.
Adjust anything that needs your judgment
If the AI guess is close but not perfect, edit the task normally. TEA is meant to reduce setup work, not replace review.
Important Behaviors
- TEA focuses on enriching newly created parent tasks rather than subtasks.
- It is designed to avoid overwriting task properties you already set yourself.
- The feature is introduced to users as
Smart Enrichmentduring onboarding. - In the currently reviewed UI, TEA does not have its own standalone settings section like ODA.
- Results depend on available context, so some tasks may only be auto-scheduled while others receive richer metadata.
Related Features
How is this guide?