> 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/context/data-context.md).

# Data context

Data context is the semantic layer over your raw data: what each column holds, its type and units, and the business names for your metrics, so the AI picks the right fields and reads their values corr

A source returns whatever its API provides: `cost_micros`, `conv_value`, `utm_source_medium`. Those column names and units come from the source app, such as Google Ads or Shopify, not from your team. Nothing in them says which of the four revenue-looking columns is the one you report on, or that a cost arrives in millionths of a currency unit.

Data context is the semantic layer over that raw data. It describes each data set and each of its columns in your own terms, so a question asked in your team's vocabulary lands on the right fields.

Coupler.io detects this layer and provides it by default, so every data set arrives with one built from its schema and its data. Your job is to refine it: the defaults describe the columns generically, and only you know the logic your data set actually holds.

## What it covers

* **What the data set represents.** The entity, the grain, and the period it covers.
* **What each column holds.** A plain-language description of the field, and what a typical value means.
* **Types, units, and allowed values.** Currency and whether an amount is in cents or micros, percentages stored as `0.15` or `15`, time zones on dates, and the full set of values a status column can take.
* **Metrics and dimensions in business language.** Your definition of ROAS, qualified lead, or active account, mapped to the columns that produce it.
* **Naming conventions and synonyms.** The internal term your team uses for a field, so the AI recognizes it when you ask.

## Refine it in Coupler.io

Open a data flow, go to the data set preview, and open a column's settings to see and edit its **description**.

{% stepper %}
{% step %}
**Review what was detected.** Coupler.io infers from column names, types, and sample values, so it gets structure right more often than meaning. Read the descriptions on the columns you actually query.
{% endstep %}

{% step %}
**Edit what doesn't match.** Rewrite the descriptions that are generic or wrong, especially units, calculated fields, and columns whose names are ambiguous. Add your team's term for the field so the AI recognizes it when you ask.
{% endstep %}

{% step %}
**Regenerate with AI when the data changes.** You can have the AI redraft descriptions for a whole data set in one step, then correct the few it got wrong rather than writing every field by hand.
{% endstep %}
{% endstepper %}

## Refine it by chatting

You can also ask your AI tool to improve the column labels and descriptions for a data set. It reads the data, proposes clearer wording, and saves it back to the data set, with no need to open Coupler.io.

A request as short as "document this data set" is enough. There's a built-in [skill](/ai/context/skills.md) for the job, so the AI follows a proper procedure instead of improvising one.

## How the AI uses it

Before running any query, the AI reads the data set's schema: column names, types, descriptions, and the context saved on it. That's what tells it which column to use for spend, that the amount is in micros, and that `status = 3` means closed-won.

Because the descriptions are stored with the data set, they apply everywhere: the AI Agent, every AI tool you've connected, and every future conversation. Update them when your data changes and the update is picked up automatically.

{% hint style="info" %}
Data context tells the AI what a column **is**. For the rules and caveats around interpreting the numbers, such as how a metric is defined, which rows to exclude, or what happened in the business last month, add [business context](/ai/context/business-context.md).
{% endhint %}

## Available in

Data context is used by the Coupler AI Agent and across every AI destination: Claude, ChatGPT, Cursor, Custom MCP, Gemini CLI, OpenClaw, and Perplexity.


---

# 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 by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.coupler.io/ai/context/data-context.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

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.