---
title: "Grafana Labs' dictionary | Writers' Toolkit documentation"
description: "Learn about the Grafana Labs dictionary template format."
---

> For a curated documentation index, see [llms.txt](/llms.txt). For the complete documentation index, see [llms-full.txt](/llms-full.txt).

# Grafana Labs’ dictionary

The Grafana Labs documentation team maintains a dictionary used for spell checking. The same dictionary is used to generate some Vale rules from the metadata in the word definition.

The dictionary uses the [Hunspell format](https://github.com/hunspell/hunspell) generated from a [Jsonnet](https://jsonnet.org) template.

If a word doesn’t yet exist in the dictionary, you can add one by modifying the appropriate [dictionary template file](https://github.com/grafana/writers-toolkit/blob/main/vale/dictionary).

To add a new word to the dictionary, refer to [Add words to the Grafana Labs dictionary](/docs/writers-toolkit/review/lint-prose/dictionary/add-words/).

## Word metadata

Each word in the dictionary has metadata that describes it. The structure of that metadata is a Jsonnet object with the following fields:

Expand table

| Key                        | Value type | Description                                                                                                                            |
|----------------------------|------------|----------------------------------------------------------------------------------------------------------------------------------------|
| `word`                     | `string`   | The spelling of the word.                                                                                                              |
| `affixes`                  | `string`   | A concatenation of the letters corresponding to Hunspell affixes. For more information, refer to [Hunspell affixes](#hunspell-affixes) |
| `po`                       | `string`   | The part of speech descriptor for the word. Known acceptable values include `'noun'`, `'adjective'`, and `'verb'`.                     |
| `description`              | `string`   | A description of the word.                                                                                                             |
| `abbreviation`             | `boolean`  | If `true`, the word is an abbreviation. Defaults to `false`.                                                                           |
| `established_abbreviation` | `boolean`  | If `true`, the abbreviation doesn’t generally need explaining, like the abbreviation HTTP. Defaults to `false`                         |
| `product`                  | `boolean`  | If `true`, the word is the name of any product, like Mimir or Facebook. Defaults to `false`.                                           |
| `Amazon`                   | `boolean`  | If `true`, the word is the name of an Amazon product, like Amazon CloudWatch. Defaults to `false`.                                     |
| `Apache`                   | `boolean`  | If `true`, the word is the name of an Apache project, like Apache Mesos. Defaults to `false`.                                          |
| `Palantir`                 | `boolean`  | If `true`, the word is the name of a Palantir product, like Apollo. Defaults to `false`.                                               |

You don’t create the object directly in the dictionary template file. Instead, you use the `word.new` function which prescribes the required fields.

To add new word, refer to [Add words to the Grafana Labs dictionary](/docs/writers-toolkit/review/lint-prose/dictionary/add-words/).

To understand the Hunspell affixes, refer to [Hunspell affixes](#hunspell-affixes).

## Hunspell affixes

Hunspell affixes teach the dictionary multiple words by applying affix rules to the stem word. The affix rules are defined in [`en_US-grafana.aff`](https://github.com/grafana/writers-toolkit/blob/main/vale/Grafana/styles/config/dictionaries/en_US-grafana.aff).

> Note
> 
> Affixes are case sensitive. `d` is a different affix to `D`

Expand table

| Letter | Kind   | Description                                                                   |
|--------|--------|-------------------------------------------------------------------------------|
| `d`    | Prefix | Adds the `de` prefix to the stem, making deprovision from provision.          |
| `m`    | Prefix | Adds the `mis` prefix to the stem, making misconfigure from configure.        |
| `p`    | Prefix | Adds the `pre` prefix to the stem, making preconfigure from configure.        |
| `u`    | Prefix | Adds the `un` prefix to the stem, making unregister from register.            |
| `D`    | Suffix | Adds the past tense suffix `ed` to the stem, making ingested from ingest.     |
| `G`    | Suffix | Adds the gerund suffix `ing` to the stem, making singing from sing.           |
| `M`    | Suffix | Adds the possessive suffix `'s` to the stem, making ingester’s from ingester. |
| `R`    | Suffix | Adds the `r` suffix to the stem, making profiler from profile.                |
| `S`    | Suffix | Adds the plural suffix `s` to the stem, making namespaces from namespace.     |