> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mibyanai.com/llms.txt
> Use this file to discover all available pages before exploring further.

# How Mibyan works in practice

> A complete guide to the Mibyan workspace, agents, context, review, and export

Mibyan is a work platform, not only a chat window. It turns an outcome you describe into a reviewable artifact or action, while keeping the conversation, files, instructions, and project context together.

## The four layers

| Layer            | What it does                                              | Where you use it        |
| ---------------- | --------------------------------------------------------- | ----------------------- |
| Mibyan app       | Chat, agents, files, artifacts, approvals, and exports    | `mibyanai.com`          |
| Workspace        | Projects, knowledge, memory, members, and settings        | Inside the app          |
| Mibyan Platform  | API keys, environments, access, budgets, usage, and logs  | `platform.mibyanai.com` |
| Mibyan model/API | Text generation and structured responses for applications | `api.mibyanai.com/v1`   |

## A typical task lifecycle

1. **Frame the outcome.** State what must exist at the end, who will use it, language, market, format, and deadline.
2. **Supply context.** Attach source files, explain definitions, and point to the project knowledge that should be trusted.
3. **Plan and execute.** Mibyan can split complex work across specialist agents and tools. Review the plan before allowing consequential actions.
4. **Inspect the artifact.** Check facts, numbers, citations, structure, language, and visual presentation in the preview.
5. **Iterate.** Give local feedback such as “replace section 3 with a comparison table” instead of restarting the task.
6. **Export or share.** Export only after checking the rendered file and removing internal notes or sensitive material.

## Choosing the right mode

* Use **chat** for brainstorming, explanations, rewriting, and decisions.
* Use **agents** for multi-step work that needs planning, tools, or a specialist perspective.
* Use **knowledge and memory** when answers must follow a durable company context or style.
* Use **artifacts** when the result should be a document, deck, spreadsheet, webpage, or other editable output.
* Use the **API** when your own product needs Mibyan generation behind a server-side integration.

## Prompt structure that works

```text theme={null}
Outcome: what should be delivered?
Audience: who will read or use it?
Context: what facts, files, or policies are authoritative?
Constraints: language, tone, length, format, deadline, and exclusions.
Acceptance criteria: how will we decide that it is finished?
```

<Warning>AI output is a draft until a person verifies important facts, calculations, legal claims, and external actions.</Warning>

## Teams and governance

Use separate projects for separate applications or environments. Keep production keys on your server, give each integration its own key, set expiry and model permissions where available, and use usage logs to investigate unexpected traffic. See [projects and members](/platform/projects-and-members), [limits and security](/platform/limits-and-security), and [usage and logs](/platform/usage-and-logs).