---
title: "Start here"
description: "What Assay is, what it is not, and where to go next."
---

<Card title="For agents" icon="robot" href="https://docs.assay.solutions/llms.txt">
  Point any agent at **https://docs.assay.solutions/llms.txt**. That file is the table of contents. The agent loads only the pages the task needs.
</Card>

Assay is a **ledger for AI work**. It records what was spent, what was said, and what the work was for — then it joins those three so you can read them together.

It runs **on your machine**, against a store you own. It is not in the request path. It cannot slow a turn and cannot break one. It reads what already happened.

Current version: **0.48.0**.

## What you can do

- Capture sessions from Claude Code, Grok, Codex, and OpenCode.
- Meter a custom agent loop as it runs.
- Open a local console: spend, sessions, products, objectives, waste, ROI, search.
- Label sittings so spend rolls up to the thing you were building.
- Share one session as a flat page — nothing else is reachable from the link.
- Search what was actually said, without sending the corpus to a model.

## What it is not

- Not a cloud dashboard that ships your prompts away.
- Not a proxy. Assay never sits between you and the model.
- Not a value oracle. It records labels with provenance. It does not decide whether the work was worth it for you.
- Not a billing page. A provider total is a sum. Assay keeps phases, attribution, and a trail for every number.

## The three records it joins

| Record | What it is |
|---|---|
| **Span** | One unit of metered work: tokens by phase, cost, who directed it. |
| **Turn** | One line of the conversation, sealed at rest. |
| **Sitting** | One harness session. This is the spine everything else hangs on. |

Cost lives on the span. Words live on the turn. Labels hang on the sitting. Assay joins them so a dollar has a conversation next to it.

## Two ways people use it

1. **Your own work with AI** — capture the coding harness you already run, then read spend and conversations in the console. Labelling is part of this path.
2. **Observability in an agent loop** — a product records each model call as it happens. Labelling is optional. The debug read is `assay turns`.

See [Applications](/applications).

## Next

1. [Install](/install) — invite, binary, `assay init`.
2. [After install](/first-run) — capture, console, sweep, labels.
3. [How it works](/how-it-works) — the data flow.
4. Agents: paste [https://docs.assay.solutions/llms.txt](https://docs.assay.solutions/llms.txt) into the agent. Human page: [For agents](/agents).