> For the complete documentation index, see [llms.txt](https://docs.goassay.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.goassay.ai/start-here/what-is-assay.md).

# What is Assay

Understand what Assay does for people using tools and API providers, and learn the first words you will meet.

Assay sits between your AI and the paid APIs it uses. It checks what each call returns, so you are not charged for broken results. Providers back the quality of their paid tools with stake.

```mermaid
flowchart LR
    A[Your AI] -->|calls a tool| B[Assay checks the result]
    B -->|sends the request| C[Provider API]
    C -->|answer| B
    D[Provider stake] -.->|backs the promise| C
```

## Pay for useful results

Before your AI pays for a tool, you can read its description, price and provider quality information. For a paid call, Assay reserves the price, checks the answer and then charges it. If a check finds that the provider returned a broken result, you are not charged, or the payment goes back to your Assay balance.

{% hint style="info" %}
**What checks can and cannot do**

* Checks vary by tool: format, required information, freshness or relevance. The receipt shows which ones ran.
* Checks do not prove that an answer is true.
* An honest empty answer, such as a search with no matches, is still charged.
  {% endhint %}

## Providers back their promise

Providers earn from the calls they serve. They set a price and a promise about how often their checked results may fail, then put up stake to back paid tools. Repeated failures beyond that promise can lead to a penalty against the stake. A refund for one call and a stake penalty are separate decisions.

## One connection for your AI

Connect your AI app to Assay once, then start with a free call. For paid tools, use a local app with assay-mcp, add funds and approve spending limits in your wallet. Hosted connectors (claude.ai, ChatGPT, and connectors added in Claude Desktop) make free calls only.

Payments use **USDG**, a dollar token, on **Robinhood Chain**.

## Words to know

| Word         | In plain words                                                                                                         |
| ------------ | ---------------------------------------------------------------------------------------------------------------------- |
| **Agent**    | The identity Assay uses to group your AI's calls and apply its rules.                                                  |
| **Tool**     | A named action your AI can ask for, with defined inputs and a result. Using it once is a call.                         |
| **Provider** | The person or business running the API (also called the endpoint) that supplies a tool's answer.                       |
| **Escrow**   | Money you deposit into a payment contract for approved tool calls. It is separate from the money still in your wallet. |
| **Stake**    | A provider's money set aside to back the quality promise for paid tools. Also called a **bond**.                       |
| **Hold**     | Money reserved for a call, so it cannot be spent twice.                                                                |
| **Recourse** | Not being charged, or getting a refund, when a call fails.                                                             |

More words are in the [Glossary](/help/glossary.md).


---

# 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.goassay.ai/start-here/what-is-assay.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.