# Accessibility audit record template

Create a dated record for each audit, named `accessibility-audit-record-<date>.md`, and retain it with the release evidence. Publish only public-safe information required by the accessibility statement.

Copy this file. Fill every cell. Do not delete a section. A blank cell is a result nobody recorded, so write `Not tested` instead.

## Results

Every result cell holds one of these four words.

| Result           | Meaning                                                                       |
| ---------------- | ----------------------------------------------------------------------------- |
| `Pass`           | Someone ran the check in every environment the row names and found nothing.   |
| `Fail`           | The check found a barrier. Link the issue or the change that fixed it.        |
| `Not tested`     | Nobody ran the check. Give the reason. It counts against a conformance claim. |
| `Not applicable` | The surface has no such content or step. Explain why.                         |

`Not tested` and `Not applicable` are different answers. The first says the evidence is missing. The second says there is nothing to gather. If the tester, assistive technology, or test data was missing, record a check that could not run as `Not tested`.

When a result differs between environments, or ran in only some of them, name the environment in the cell, for example `Fail (E2)` or `Pass (E1)`.

## Audit

| Field                                  | Value |
| -------------------------------------- | ----- |
| Dates                                  |       |
| Revision and release                   |       |
| Scope and sample rationale             |       |
| Deployment and test data               |       |
| Evaluator                              |       |
| WCAG-EM report                         |       |
| Storybook report, revision and browser |       |
| Finding query                          |       |

## Environments

Give each environment an ID (`E1`, `E2`, …) and list what was run in it. An environment that was planned and not run stays in the table as `Not tested`.

| ID | Operating system | Browser | Assistive technology | Versions and relevant settings | Run |
| -- | ---------------- | ------- | -------------------- | ------------------------------ | --- |
|    |                  |         |                      |                                |     |

## Findings

One row per failure, with the issue or change that fixes it and the story or test that fails without the fix.

| Finding | Surface | Criterion | Barrier | Fix or issue | Coverage |
| ------- | ------- | --------- | ------- | ------------ | -------- |
|         |         |           |         |              |          |

## Surfaces

`Automated` is the axe run, plus the title, reflow and motion checks. `Keyboard` is a Tab walk of the surface: order, focus indicator, focus not hidden and no trap. Link each failure to its issue.

| Surface | Route, state and role | Automated | Keyboard | NVDA | VoiceOver | Evidence or issue |
| ------- | --------------------- | --------- | -------- | ---- | --------- | ----------------- |
|         |                       |           |          |      |           |                   |

## Complete processes

One row for each process in each environment. A process is complete when a person finishes it from the first step to the last, including the error they make on the way.

| Process | Roles and states | Environment | Success path | Error recovery | Evidence or issue |
| ------- | ---------------- | ----------- | ------------ | -------------- | ----------------- |
|         |                  |             |              |                |                   |

## Remaining work

List each `Not tested` result and what a tester has to run to close it.