> For the complete documentation index, see [llms.txt](https://docs.carto.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.carto.com/carto-for-agents/carto-for-agents.md).
# Overview
**CARTO for Agents** brings the full CARTO platform to your AI assistant. Connect your favorite agent — Claude, ChatGPT, Gemini Enterprise, Microsoft Copilot Studio, and more — and create and edit maps, build and run workflows, explore your data warehouse, and manage your organization, all in natural language.
Most teams get everything they need from two pieces:
* [**CARTO MCP Server**](/carto-for-agents/mcp-server.md) — a hosted [Model Context Protocol](https://modelcontextprotocol.io/) server that plugs CARTO into any MCP-compatible AI client. Your agent lists connections, browses data, creates and edits Builder maps, authors and runs Workflows, renders maps inline, and reaches admin tools — over a single URL. This is the starting point for most users.
* [**CARTO Agent Skills**](/carto-for-agents/agent-skills.md) — a public catalog of playbooks at [`CartoDB/agent-skills`](https://github.com/CartoDB/agent-skills) that teach coding agents (Claude Code, Codex, Gemini CLI, and others) the right patterns for driving CARTO, so they get it right the first time.
For scripting, automation, and deeper administration, there's a third piece:
* [**CARTO CLI**](/carto-for-agents/cli.md) — a command-line interface (`@carto/carto-cli`) that exposes nearly every platform operation as a script-friendly command. Ideal for pipelines and headless environments, for bulk and cross-organization operations, and for administration and auditing — including SQL over your usage logs. The most complete surface, for power users who want fine-grained control.
## Where to start
| You want to… | Use this |
| --------------------------------------------------------------------------------------------------------------------- | -------------------------------------------------------------------------------- |
| Set CARTO up in your AI platform — Claude, ChatGPT, Gemini, Copilot, Snowflake, Databricks, coding agents, and more | [Connect your platform](/carto-for-agents/connect-your-platform.md) |
| Understand the in-conversation tools — create maps, build workflows, explore data, and more — and how to authenticate | [CARTO MCP Server](/carto-for-agents/mcp-server.md) |
| Give a coding agent (Claude Code, Codex, Gemini CLI) the right CARTO patterns | [CARTO Agent Skills](/carto-for-agents/agent-skills.md) |
| Script CARTO from a terminal or CI/CD pipeline, run bulk or cross-org operations, or audit usage logs | [CARTO CLI](/carto-for-agents/cli.md) |
| Expose your own analytical workflow as a tool an agent can call | [Workflows as MCP Tools](/carto-user-manual/workflows/workflows-as-mcp-tools.md) |
{% hint style="success" %}
**New here?** The fastest path is the [CARTO for Agents quickstart](/getting-started/quickstart-guides/connecting-carto-to-your-agents.md): connect an agent to the MCP Server and paste one prompt to build your first map.
{% endhint %}
## How the pieces fit together
The three surfaces are complementary, not competing. Pick the one that matches how you work — or combine them.
* The **MCP Server** is the shortest path from an AI client to CARTO. It needs no install: you connect a URL, sign in, and your agent can immediately reach a broad set of platform tools. Best for chat clients (Claude, ChatGPT, Gemini, Copilot, Databricks, Snowflake) and any MCP-compatible host.
* **Agent Skills** ride on top of coding agents that run on your machine. They don't replace the MCP Server or the CLI — they make the agent *use them well*, choosing the right approach, flags, SQL dialect, and job-handling pattern for each request.
* The **CLI** runs locally and can be called by any agent that executes shell commands, or by you directly. It's the most complete surface — the right tool for automation and scripting, for bulk and cross-organization operations, and for administration and auditing, including querying usage logs with SQL.
Workflows you build in CARTO can be published as new tools. The same published workflow is reachable from the MCP Server, referenced by the Skills, and callable from the CLI.
{% hint style="info" %}
Looking for the **AI Agents you build inside Builder maps** (in-product agents bound to a map's data and tools)? That's a different product surface — see [CARTO User Manual → AI Agents](/carto-user-manual/ai-agents.md). Want to **embed AI-powered map interaction in your own application**? See [CARTO for Developers → Build an AI-powered map application](/carto-for-developers/guides/build-an-ai-powered-map-application.md).
{% endhint %}
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:
```
GET https://docs.carto.com/carto-for-agents/carto-for-agents.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
