> For the complete documentation index, see [llms.txt](https://docs.coupler.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.coupler.io/ai/mcp.md). # MCP MCP (Model Context Protocol) is an open standard for connecting AI tools to outside systems. Coupler.io runs an MCP server, so once you connect it, your AI tool can work with your data and your data flows directly — no exports, no copy-paste, no API keys to manage. You connect once per AI tool. After that, the AI has a set of **tools** it can call on your behalf, listed below. You never call these yourself: you ask a question in plain language, and the AI picks the tools it needs. ## Which data sets are visible An AI tool connected over MCP sees only the data sets from data flows that have **that tool** as a destination. Adding Claude as a destination makes a data flow available in Claude — it does not make it available in ChatGPT. To use the same data from several tools, add each one as a destination. The Coupler.io AI Agent is different: it runs inside Coupler.io and can see every data set in your workspace, with no destination step needed. {% hint style="info" %} If your AI tool can't find a data set you expect, the usual reason is that its data flow hasn't been sent to that tool yet. Add the tool as a destination on that data flow. {% endhint %} ## What the AI can and can't do **It can read your data.** Queries are read-only — the AI cannot edit, delete, or overwrite the rows Coupler.io has imported. **It can change your configuration.** With the right tools available, the AI can create data flows, add sources and destinations, change a schedule, rename columns, build SQL data sets, and trigger a run. Those are real changes to your workspace, so the AI is instructed to confirm with you before making one you didn't ask for. **It sees nothing else.** The AI reaches the data sets described above and nothing else — not your other accounts, not your machine. ## Finding your data | Tool | What it does | | ----------------- | ---------------------------------------------------------------------------------------------------------------------- | | `list-datasets` | Lists the data sets across your data flows. This is how the AI answers "what data do I have?" | | `search-datasets` | Finds a data set by name, source, or keyword — faster than listing everything when you already know what you're after. | | `list-templates` | Browses Coupler.io's ready-made templates, so the AI can suggest a pre-built setup instead of assembling one. | ## Reading and analyzing | Tool | What it does | | ------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `get-schema` | Reads a data set's columns — names, types, descriptions, and any business context saved on it. The AI reads this before querying, so it interprets your columns correctly. | | `get-data` | Runs a SQL query against a data set and returns the result. This is what actually answers your question. The querying happens on Coupler.io's side, so large data sets don't have to fit into the AI's context. | ## Shaping and combining data | Tool | What it does | | ----------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `create-dataset` | Builds a new SQL data set from data sets in the same data flow — a join, an aggregation, a pivot. The AI uses this when a question needs two data sets combined, since one query reads one data set at a time. See [SQL based transformations](https://docs.coupler.io/functionality/data-set/combining-data/sql-based-transformations). | | `update-dataset-query` | Changes the SQL behind an existing SQL data set. | | `update-dataset-schema` | Improves column labels and descriptions, so both you and the AI can read the data set more easily. | ## Building and configuring data flows | Tool | What it does | | ------------------------------------------------------------- | ------------------------------------------------------------------------------ | | `create-dataflow` | Creates an empty data flow to attach a source and destination to. | | `create-dataflow-from-template` | Creates a data flow from a template, already configured. | | `get-dataflow` | Reports a data flow's current state — its sources, destinations, and last run. | | `create-dataflow-source` / `update-dataflow-source` | Adds or changes the source data is pulled from. | | `create-dataflow-destination` / `update-dataflow-destination` | Adds or changes where the data is delivered. | | `update-dataflow` | Renames a data flow, or changes and pauses its schedule. | ## Connecting to services | Tool | What it does | | ------------------------------- | --------------------------------------------------------------------------------------------------------------- | | `list-credentials` | Checks which services you're already authorized for, and gives you a link to connect a new one. | | `list-integrations` | Lists the services Coupler.io supports as a source or a destination. | | `get-integration` | Reads what a specific integration needs — its parameters and how it authenticates. | | `get-integration-field-options` | Fills in the choices for a setting that depends on your account, such as which ad account or which spreadsheet. | ## Remembering and refreshing | Tool | What it does | | ---------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | `update-dataset` | Saves what the AI learned about a data set — what a metric means, which currency it's in, which rows to exclude — so the next conversation starts informed. See [Dataset context](/ai/dataset-context.md). | | `run-dataflow` | Refreshes a data flow from its source. The AI only does this when you ask. | ## Skills | Tool | What it does | | ------------- | ------------------------------------------------------------------------------------------------ | | `list-skills` | Checks which skills are available for your workspace. | | `get-skill` | Loads a skill's instructions, so the AI follows a proven procedure. See [Skills](/ai/skills.md). | ## Available in MCP powers every Coupler.io AI destination: Claude, ChatGPT, Cursor, Custom MCP, Gemini CLI, OpenClaw, and Perplexity. For how to connect a specific tool, see that tool's page under [Destinations → AI](https://docs.coupler.io/destinations/categories/ai). --- # 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.coupler.io/ai/mcp.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.