> ## Documentation Index
> Fetch the complete documentation index at: https://docs.knify.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Pixel audit demo

> A concrete early demo for interactive AI reports.

A strong first demo is a PixelFix audit.

```text theme={null}
PixelFix audit
-> agent inspects pixel events
-> agent outputs Knify canvas
-> client opens share link
-> downstream system reads exported issue list
```

## Canvas contents

The report should include:

* Summary
* Issue severity
* Charts
* Event logs
* Detected misfires
* Consent problems
* Recommended fixes
* Rerun audit button
* Share-with-client link
* Exported issue list API

## Why this wedge works

It is concrete, commercial, and visually obvious.

It shows Knify is not only a dashboard builder because the artifact is agent-native, patchable, linkable, forkable, API-exportable, and runtime-agnostic.

## Example SDK call

```ts theme={null}
import { Knify } from "@knify/client";

const knify = new Knify({ token: process.env.KNIFY_TOKEN });

await knify.canvases.create({
  title: "Pixel Attribution Audit",
  kind: "report",
  blocks: [
    {
      id: "summary",
      kind: "text.markdown",
      props: {
        markdown: "Found 317 misfires and 42 missing-consent events."
      }
    }
  ],
  exports: [
    {
      name: "misfire_count",
      path: "/data/summary_metrics/misfires",
      type: "number",
      stable: true
    }
  ]
});
```