Build the agent

Brief an agent in plain language

Turn your team’s instructions into an agent.

Captured from the product Demo workspace
Describe the job and its rules in plain language.

What this changes for your team.

Procedure describes the job, and Principles set the rules the agent should follow. Your team can write both in everyday language and review the agent’s responses and actions in its session history.

How it works in practice.

  1. 01

    Write the ordered job in Procedure, using the same language your team uses in a handoff.

  2. 02

    Add non-negotiable boundaries and decision rules under Principles.

  3. 03

    Test the draft, read what it did in the session's Activity tab, and tighten the instructions before publishing.

What you can plan around.

What to expect, including limits that matter to your team.

An agent draft stores Procedure and Principles as first-class setup fields, not as a generated graph.

Published sessions pin the exact agent version they execute.

You can review the agent’s messages and tool use separately in its history.

Bring one real process

See how Yekar.AI fits the way you work.

Start with a job your team already owns, plus the tools and decisions around it.

Get started