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

# Docs for AI agents

> Give your AI assistant or coding agent these docs, Minitap's tools and the CLI's skill, so it answers and acts with the current product.

Your coding agent or AI assistant can read these docs and drive Minitap for you. This page lists every entry point an agent can use, from a single page in Markdown to the tools that create scenarios and start runs.

| Your agent needs to | Give it |
| - | - |
| Answer questions about Minitap | The [docs MCP server](#search-the-docs-from-your-agent) or [llms.txt](#plain-text-for-any-model) |
| Read one page | The page URL with `.md` at the end, or **Copy page** at the top of the page |
| Write scenarios, start runs, read results | The [Minitap MCP server](https://www.minitap.ai/docs/minitest/integrations/cursor-and-claude.md#connect-the-mcp-server) |
| Script Minitap from a shell or CI | The [`minitest` CLI](#the-cli-and-its-skill), with its agent skill |

## Plain text for any model

Every page of these docs exists as Markdown, and the whole site has an index built for language models.

| URL | What it holds |
| - | - |
| [`/docs/llms.txt`](https://www.minitap.ai/docs/llms.txt) | One line per page: title, link, one-sentence summary |
| [`/docs/llms-full.txt`](https://www.minitap.ai/docs/llms-full.txt) | Every page in full, in one file |
| Any page URL + `.md` | That page in Markdown, for example [`/docs/minitest/get-started/quickstart.md`](https://www.minitap.ai/docs/minitest/get-started/quickstart.md) |

The menu at the top of each page also copies the page as Markdown, or opens it in ChatGPT or Claude with the page attached.

## Search the docs from your agent

The docs have their own MCP server at `https://www.minitap.ai/docs/mcp`. It needs no sign-in. Your agent can search the docs, read any page in full, and report a page that is wrong or out of date.

> [tab] Claude Code
>
> ```bash theme={"theme":{"light":"github-light","dark":"vesper"}}
> claude mcp add --transport http minitap-docs https://www.minitap.ai/docs/mcp
> ```

> [tab] Cursor
>
> In `.cursor/mcp.json`, or `~/.cursor/mcp.json` for every project:
>
> ```json theme={"theme":{"light":"github-light","dark":"vesper"}}
> {
>   "mcpServers": {
>     "minitap-docs": {
>       "url": "https://www.minitap.ai/docs/mcp"
>     }
>   }
> }
> ```

> [tab] Other clients
>
> Add a remote HTTP server with the URL `https://www.minitap.ai/docs/mcp`.

Then ask your agent, for example: "With the Minitap docs, explain how to sign Mini in with a phone number on our staging backend."

## Act on your workspace

The docs server reads documentation. To let your agent work in your workspace, add the Minitap MCP server too: it signs in with your Minitap account and can list apps, write and edit scenarios, manage test accounts, start runs and read their verdicts. Setup for Claude Code, Cursor and other clients is in [Coding agents](https://www.minitap.ai/docs/minitest/integrations/cursor-and-claude.md). Both servers can run side by side.

## The CLI and its skill

The `minitest` CLI ships with an agent skill: the guide that teaches a coding agent the commands, the rules for good scenarios and how to read results. It stays in step with the CLI version you run.

```bash theme={"theme":{"light":"github-light","dark":"vesper"}}
# Install, then sign in once per machine
curl -fsSL https://raw.githubusercontent.com/minitap-ai/minitest-cli/main/install.sh | bash
minitest auth login

# From your app's repository: the onboarding plan for your agent
minitest init

# The skill itself, and the reference documents behind it
minitest skill
minitest reference
```

`minitest upgrade` updates the CLI and refreshes the skill. Every command takes `--json`, so your agent reads structured output. The full list is in [CLI commands](https://www.minitap.ai/docs/minitest/reference/cli-commands.md).

## Tips for agents

* **Use the app's words.** Minitap calls a test a scenario, its checks acceptance criteria, and the users Mini signs in as test accounts. The [glossary](https://www.minitap.ai/docs/minitest/reference/glossary.md) lists them all.
* **Check what Mini can do first.** `minitest capabilities`, or [What Mini can do](https://www.minitap.ai/docs/minitest/reference/capabilities.md), says which sign-ins, gestures and device features each platform supports.
* **Fix from the report.** When a scenario fails, the **Fix with your agent** tab of the run report holds a prompt written for your coding agent. See [Read a run report](https://www.minitap.ai/docs/minitest/runs/run-report.md#fix-prompt).

## Next steps

> [card] [Coding agents](https://www.minitap.ai/docs/minitest/integrations/cursor-and-claude.md)
>
> Connect Cursor or Claude Code to your workspace.

> [card] [CLI commands](https://www.minitap.ai/docs/minitest/reference/cli-commands.md)
>
> Every command and flag.

> [card] [What a scenario is](https://www.minitap.ai/docs/minitest/suite/anatomy.md)
>
> The shape your agent writes.

> [card] [Glossary](https://www.minitap.ai/docs/minitest/reference/glossary.md)
>
> The words Minitap uses everywhere.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.