YOUR AI CREW / 3 MINUTES

A good first conversation

A clear request is a small kindness to your future self.

Your little win: Give an employee an actionable brief and follow its work.

01 / SEE HOW IT FITS

The big picture, in little steps.

STEP 1 OF 4

Say what done looks like

Describe the outcome, relevant context, constraints and expected output. “Draft a warm kickoff email with these three agreed milestones” is clearer than “handle the client”.

02 / YOUR TURN

Have a little go.

✎ fictional practice

This is your practice table. Everything stays here—no account, real AI calls, or changes to Seerist.

Setting out the pencils…

03 / DO IT IN SEERIST

Ready for the real thing?

AI Employees → Chat

Before you start: Relevant workspace context and any capabilities the task requires.

  1. 1

    Say what done looks like

    Describe the outcome, relevant context, constraints and expected output. “Draft a warm kickoff email with these three agreed milestones” is clearer than “handle the client”.

  2. 2

    Follow up in the same thread

    The conversation preserves context. Read progress and results; a partial answer or pending action is not finished work.

  3. 3

    Review the handoff

    Open saved artifacts or linked resources. Exact external actions can wait for approval. A tool connection must be granted before an employee can use it.

  4. 4

    Keep it organized

    Project hub → Conversations supports search, naming, pinning, archiving and folder assignment. Use a descriptive title so you can return to the decision later.

A little thing to rememberBackground work can continue after you leave. Check its status before starting another copy of the same task.

Open this area in Seerist ↗ (new tab)

Opens the real app in a new tab. Sign in and check the selected workspace before acting.

Take a little note with you.

About this lesson

Checked September 13, 2026 against app source a365066. Based on web/src/components/AgentChat.tsx; docs/connected-workspace.md. This is a source-based guide, not a claim that every provider action was live-tested.