Crew Tasks

Some work shouldn't need you at all: the morning project digest, the pre-meeting briefing, the weekly pipeline review. Crew tasks are exactly that — jobs you define once and the crew runs on a trigger, delivering the result as a note, an event description, or a notification. You manage them all in Settings → Crew Tasks.

Adding your first task

Prefer starting from scratch? Define your own task builds one from a plain-language description instead of a template.

Triggers — when tasks run

• On a schedule — daily or weekly at a time you pick, in your timezone. Digest-style tasks default to sensible slots (morning digests, Friday reviews).
• Before an event — runs a set lead time before matching calendar events; how the Meeting Briefing lands in your event description an hour before the call.
• When an email is labeled — fires on Gmail labels (label an email Fokus/Contracts and the contract review starts) or on Email Assistant triage categories (travel confirmations feed the Trip Companion automatically).
• Manually — every task supports Run now, whatever its trigger.

Command them from chat

Captain Fokus is the crew's chief, so you rarely need this settings page day-to-day:

• "Run my meeting briefing" — triggers any of your tasks by name.
• "What did my crew find?" — recent runs across all tasks, with the content summarized in chat instead of you opening notes.
• "Have the crew research …" — delegates a one-off job from the catalog. The Captain creates it disabled (it won't schedule itself), runs it once, and you can promote it to a recurring task here if it earns its keep.
• If a run pauses on a question, you can answer it right in chat.

What happens in a run

Every run follows the same shape: the task gathers from its sources (your workspace, email, the web, connectors), composes what it found into the sections you asked for, and delivers — usually a note, sometimes an event description, plus an optional notification.

Two behaviors worth knowing:

• Quiet periods are skipped. If a run gathers nothing worth reporting — no new deals, an empty week — it ends as skipped: no note, no noise.
• Partial beats nothing. If one source fails (say, one connected app is temporarily unreachable), the run continues with the rest and the run history shows exactly which source failed.

Each task keeps a run history with per-step summaries, so you can always see what a run read and what it produced.

Making a task yours

Open any task to tune it:

• Trigger & cadence — switch trigger types where allowed, change the schedule.
• Instructions — the task's editable brief. This is where the interesting customization lives: the contract review's negotiation playbook, the Trip Scout's home airport and budget, the SEO Watch's tracked domain. Edit it like a note to yourself — the crew follows it.
• Depth & sections — how thorough the composed result should be and which sections it produces.
• Skills — the individual steps of the task; each can be configured, and optional steps can be added or removed.

Custom skills

If the built-in gathering steps don't cover a source you care about, build your own under Custom skills on the same page: describe what to fetch and how, and your skill becomes available as a step in any crew task — same budget, same read-only rules.

Budget and limits

The Daily AI energy meter at the top of the page shows how much of today's AI budget each part of the crew has used — crew tasks draw from the same daily allowance as Captain Fokus. There's also a daily cap on total runs, and research-style tasks (the radars, competitive intel, job enrichment) draw on the Research add-on's monthly search allowance.

Where to next

• Browse the full Task Catalog — all 35 predefined tasks.
• Connect apps for the connector-powered tasks: Connectors.
• See what the crew knows about you and your contacts: AI Profiling.