# Docs for AI Agents

Everything in this documentation is machine-readable. If you're an AI agent — or you're pointing one at these docs to implement KinesteX — use the surfaces below instead of scraping HTML.

**Start here:** fetch [llms.txt](https://kinestex.com/llms.txt) for an index of every page with descriptions, or [llms-full.txt](https://kinestex.com/llms-full.txt) for the entire documentation as a single Markdown file.

## Read any page as Markdown

Append `.md` to any documentation URL to get the page as clean Markdown with code samples for **all platforms**:

```text
https://kinestex.com/docs/installation.md
https://kinestex.com/docs/data-points/workout-events.md
```

The response is `text/markdown` and always in sync with the page.

## Connect over MCP

The docs are exposed as a [Model Context Protocol](https://modelcontextprotocol.io) server with search and read tools at `https://kinestex.com/mcp` (Streamable HTTP transport).

```bash
# Claude Code
claude mcp add --transport http kinestex-docs https://kinestex.com/mcp
```

```json
// Cursor and other MCP clients (.mcp.json / mcp.json)
{
  "mcpServers": {
    "kinestex-docs": { "url": "https://kinestex.com/mcp" }
  }
}
```

**Available tools:**
| Tool | Description |
|------|-------------|
| search_docs | Search the documentation, returns matching pages with excerpts |
| read_doc | Read a full page as Markdown by its path (e.g. `installation` or `data-points/workout-events`) |
| list_docs | List every documentation page with its path, title, and description |

## Structured JSON API

`GET /api/docs` returns the entire documentation as structured JSON. Filter with `?sections=` (comma-separated section IDs; nested paths supported):

```text
https://kinestex.com/api/docs?sections=installation,data-points/workout-events
```

Rate limit: 50 requests per minute per IP (see `X-RateLimit-Remaining` header).

## Recommended flow for implementing KinesteX

1. Fetch [llms.txt](https://kinestex.com/llms.txt) and identify the platform you're building for (Swift, Kotlin, React Native, Flutter, HTML & JS, or React TS).
2. Read `/docs/installation.md`, `/docs/configuration.md`, and the integration option you're implementing — every code block covers all six platforms.
3. Read `/docs/data-points/receiving-data.md` and the event pages under `/docs/data-points/` to handle SDK events.
4. Check `/docs/customization-parameters/` pages for the exact parameter names your integration needs.

**Note:** an API key is required to run the SDK. A human on your team can request one via the [contact form](https://kinestex.com/#contact-form).

**Product changelog** is machine-readable too: [updates.kinestex.com/api/changelogs](https://updates.kinestex.com/api/changelogs) returns all release notes as JSON.

---
Source: https://kinestex.com/docs/ai-agents · Index: https://kinestex.com/llms.txt