What to understand
The main idea of this section: create and use custom AI agents for your team's repeatable tasks. It answers a practical user question and shows which limitations to take into account in advance.
Documentation
Custom agents
Custom agents — a practical Keter Labs reference on how to create and use custom AI agents for your team's repeatable tasks. This section helps you quickly understand the preparation, the limitations, the sequence of steps and the criteria for a finished result without lengthy searching across different pages.
docshelpguide
What's included
What this section gives you.
The core of the task without extra searching
This page explains how to create and use custom AI agents for your team's repeatable tasks. The user gets working context rather than a general overview: where to start, what is important to check and when to move on to the next step.
Preparation before you act
Before you start, it is worth preparing the agent's role, input data, brand rules, constraints, examples of a good result and access permissions. This reduces the number of random iterations and helps the team agree on the expected result in advance.
Checks before you continue
After completing a step, check the stability of responses, compliance with the rules, saved instructions, access and the quality of repeated runs. Once these points are covered, the material, setting or process can safely be handed over.
When to use it
Common jobs this section is built for.
These examples make it easier to choose where to start and which nearby section to open next.
You need to understand a feature quickly
Open this section when you need to create and use custom AI agents for your team's repeatable tasks but do not want to spend time searching for similar answers in different parts of the help center.
The team works to a single process
The material is easy to send to colleagues: it records the preparation, the expected result, the checks and the neighboring sections needed for a complete workflow.
Before publishing or handing over
Use this page as a checklist: the stability of responses, compliance with the rules, saved instructions, access and the quality of repeated runs should be clear before the task goes to a client, the team or production.
Details
What to know before you use it.
How to apply it in your work
Start with the goal and input materials, then follow the steps in the interface and save decisions in the project. For this topic, the agent's role, input data, brand rules, constraints, examples of a good result and access permissions are especially important.
Where to go next
If the task is broader than one section, move on to related topics: AI assistant, MCP, Spaces and workflow automation. This way the user keeps the context and assembles a complete workflow faster.
Workflow
How it works inside Keter Labs.
Every section in Keter Labs follows the same shape: a clear brief, a fast first draft, refinement in the editor, and a result saved into the project.
1
Step 1
Define the specific task: create and use custom AI agents for your team's repeatable tasks. Record who is responsible for the result and where it will be used.
2
Step 2
Prepare the agent's role, input data, brand rules, constraints, examples of a good result and access permissions. For team work, check roles, access and where materials are stored right away.
3
Step 3
Perform the action in Keter Labs and save important parameters, versions, files, prompts or links inside the project.
4
Step 4
Before finishing, check the stability of responses, compliance with the rules, saved instructions, access and the quality of repeated runs; then move on to related topics: AI assistant, MCP, Spaces and workflow automation.
Related sections
Where to go next.
Help center
Keter Labs help center
The "Keter Labs help center" section collects practical guidance on the topic: find the right instructions on tools, plans, permissions, credits, the API and teamwork. The section helps you prepare materials, complete the steps in Keter Labs, check the limits and bring the result to a working state.
Open
Documentation
Quickstart
The "Quickstart" section brings together practical guidance on a single topic: how to get started in Keter Labs: create a project, choose a tool, add references and save your first result. It helps you prepare materials, complete the steps in Keter Labs, check the limitations and bring the result to a working state.
Open
Documentation
Projects
The "Projects" section brings together practical guidance on a single topic: how to organize files, versions, assets and results inside Keter Labs projects. It helps you prepare materials, complete the steps in Keter Labs, check the limitations and bring the result to a working state.
Open
Documentation
Troubleshooting
The "Troubleshooting" section gathers practical guidance on one topic: quickly diagnose errors with generation, access, files, payments, exports, and result quality. It helps you prepare materials, complete the steps in Keter Labs, check the limitations, and bring the result to a working state.
OpenReady when you are
Bring this workflow into your next project.
Spin up assets, lock in versions and align your team around a single creative platform.
Create account